Advances in Physical Organic Chemistry

Advances in Physical Organic Chemistry This Page Intentionally Left Blank Advances in Physical Organic Chemistry Vo...

103 downloads 1843 Views 14MB Size Report

This content was uploaded by our users and we assume good faith they have the permission to share this book. If you own the copyright to this book and it is wrongfully on our website, we offer a simple DMCA procedure to remove your content from our site. Start by pressing the button below!
Report copyright / DMCA form

DOWNLOAD PDF

Advances in Physical Organic Chemistry

This Page Intentionally Left Blank

Advances in Physical Organic Chemistry Volume 23 Edited hj,

D. B E T H E L L The Robert Robinson Laboratories University of Liverpool P.O. Box 147, Liverpool L69 3BX

A C A D E M I C P R E S S 1987 Hurcourl B r o w Jovanovirh. Puhliskers London Orlando San Diego New York Austin Boston Sydney Tokyo Toronto

ACADEMIC PRESS INC. (LONDON) LTD 24/28 Oval Road London NWI 7DX Unired Srares Edilion published b!. ACADEMIC PRESS INC. Orlando. Florida 32887

Copyright 0 1987 by ACADEMIC PRESS INC. (LONDON) LTD

No part of this book may be reproduced in any form by photostat, microfilm, or any other means. without written permission from the publishers

ISBN 0-12-033523-9 ISSN 0065-3 I60

TYPESET BY BATH TYPESETTING LTD.. BATH, U.K. A N D PRINTED IN GREAT BRITAIN BY ST. EDMUNDSBURY PRESS, BURY ST. EDMIJNIX.

Contents Contributors t o Volume 23

vi i

The Nucleophilicity of Metal Complexes Towards Organic Molecules

1

S U N D U S H E N D E R S O N and R I C H A R D A. H E N D E R S O N 1

2 3 4 5 6 7 8 9 10 11

12 13 14 15

16 17

The scope of the review 2 Introduction to metal complexes as nucleophiles 2 Types of metal-based nucleophiles 5 The scale of nucleophilicity 6 The influence of the solvent on the nucleophilicity 14 Stereochemical changes at a saturated carbon centre I8 Stereochemical changes at an unsaturated carbon centre 30 Stereochemical changes at the metal 32 The iodide catalysis effect 38 The reactions of binuclear complexes 39 The reactivity of the carbon centre 41 The reactions of am-dihalogenoalkanes 44 Activation parameters 47 Thermodynamics of reactions involving metal nucleophiles 49 Activation of carbon-hydrogen bonds 50 Applications 53 Ad,finem 58

Isotope Effects on nmr Spectra of Equilibrating Systems

HANS-ULLRICH SlEHL 1 Introduction 2 Applications

63 82 V

63

vi The Mechanisms of Reactions of p- Lacta m Antibiotics

165

M I C H A E L I. P A G E 1 Introduction 166 2 Mode of action of p-lactam antibiotics 173 3 Is the antibiotic p-lactam unusual? 184 4 Alkaline hydrolysis and structure: chemical reactivity and relationships 198 5 Acid hydrolysis 207 6 Spontaneous hydrolysis 2 15 7 Buffer catalysed hydrolysis 2 16 8 Metal-ion catalysed hydrolysis 218 9 Micelle catalysed hydrolysis of penicillins 223 10 Cycloheptaamylose catalysed hydrolysis 232 1 1 The aminolysis of p-lactam antibiotics 233 12 The stepwise mechanism for expulsion of C(3’)-leaving groups in cephalosporins 250 13 Reaction with alcohols and other oxygen nucleophiles 252 14 Epimerisation of penicillin derivatives 258

Free Radical Chain Processes in Aliphatic Systems involving an Electron Transfer Process

271

G L E N A. R U S S E L L Introduction 271 Free radical chain processes Free radical chain processes between neutral substances 4 Free radical chain reactions 5 Concluding remarks 3 15 I 2 3

involving nucleophiles 274 involving electron transfer 299 involving radical cations 308

Author Index

323

Cumulative Index of Authors

340

Cumulative Index of Titles

342

Contributors to Volume 23 Richard A. Henderson AFRC Unit of Nitrogen Fixation, The University of Sussex, Brighton, BNI 9RQ, U.K. Sundus Henderson A F R C Unit of Nitrogen Fixation, The University of Sussex, Brighton, BNl 9RQ, U.K. Michael I. Page Department of Chemical and Physical Sciences, The Polytechnic, Queensgate, Huddersfield, H D 1 3DH, U.K. Hans-Ullrich Siehl Institute of Organic Chemistry, The University of Tubingen, D-7400 Tubingen I , Auf der Morgenstelle 18, Germany Glen A. Russell Department of Chemistry, Iowa State University, Ames, Iowa 50011, U.S.A.

vii


The Nucleophilicity of Metal Complexes Towards Organic Molecules S U N D U S H E N D E R S O N a n d R I C H A R D A. H E N D E R S O N A.F.R.C. Unit of Nitrogen Fixation, University of Sussex, Brighton BN1 9RQ, U.K.

5

6

7

8 9 10

II 12 13 14 15 16 17

The scope of the review 2 Introduction to metal complexes as nucleophiles 2 Types of metal-based nucleophiles 5 The scale of nucleophilicity 6 The influence of ligands on the nucleophilicity 7 The nucleophilicity of metal complexes 12 The influence of the solvent on the nucleophilicity 14 Stereochemical changes at a saturated carbon centre 18 Reaction of coordinatively-saturated complexes 18 Reactions of coordinatively-unsaturatedcomplexes 22 Reactions of Me,Sn- 27 Reactions of binuclear complexes 29 Stereochemical changes at an unsaturated carbon centre 30 Reactions of coordinatively-saturated complexes 30 Reactions of coordinatively-unsaturated complexes 3 I Stereochemical changes at the metal 32 The iodide catalysis effect 38 The reactions of binuclear complexes 39 The reactivity of the carbon centre 41 The reactions of a,w-dihalogenoalkanes 44 Activation parameters 47 Thermodynamics of reactions involving metal nucleophiles 49 Activation of carbon-hydrogen bonds 50 Applications 53 Ad,finmz 58 References 58

ADVANCES IN PHYSICAL ORGANIC CHEMISTRY VOLUME 23

ISBN

n

1 2 033523 Y

1

Copvrt&dir D I987 Acrrdmm P r w lm London

All righis o/reproducrion in anv lomi r~scvrcd

2

SUNDUS HENDERSON AND RICHARD A. HENDERSON

1 The scope of t h e review

Nucleophilic aliphatic substitution, as defined by ( I ) , and typified in organic chemistry by the Finkelstein and Menschutkin reactions, can extend into inorganic chemistry with the inclusion of metal complexes which are sufficiently nucleophilic. The aim of this review is to collect the somewhat Y-

+ R-X-R-Y

+ X-

(1)

diverse information which exists on the nucleophilicity of certain metal complexes towards organic molecules. Although this class of reaction has often been included in previous reviews, the main emphasis of these articles has been to compare the nucleophilic and redox pathways of these complexes, rather than to give a detailed discussion of the nucleophilicity. This review will concentrate on the detailed mechanistic study of the reactions between metal-complex nucleophiles and organic molecules including the stereochemical consequences of the reactions at both the carbon centre (Sections 6 and 7) and the metal (Sections 8 and 9). Furthermore, the influence of substituents on the carbon centre (Sections 1 1 and 12) and the nature of the solvent (Section 5) will be discussed. However, neither the nucleophilicity of ligands towards organic substrates nor the synthesis of nucleophilic complexes will be covered. The reader interested in the latter is recommended to the reviews by King (1975) and Vaska (1968). 2

Introduction t o metal complexes as nucleophiles

The idea of a simple metal complex acting as a nucleophile is conceptually relatively simple. All that is required is a metal centre (the metal and its coligands) which is relatively electron-rich, and which has a vacant site, or a potentially vacant site (via rapid dissociation of a Iigand). This is typified by the reaction shown in (2) between the five-coordinate, square-based pyramidal [Co(dmgH),py] (py = pyridine, dmgH = dimethylglyoxime) and RX. [Co(dmgH),pyl-

+ R-X

-

[Co(R)py(dmgH),]

+ X-

(2)

The analogy between ( I ) and (2) is obvious. However, if the metal nucleophile contains more than one vacant site the liberated group X - can bind to the metal. as shown in the archetypal reaction of this type shown in (3), involving the square-planar trans-[IrCI(CO)(PPh3),].

+

~rrrri.s-[lrCl(CO)(PPh,)zl Me1

-

[Ir(Cl)(I)Me(CO)(PPh,),]

(3)

I t is worthwhile discussing the salient features of reaction (3) at this stage, since they will encompass many of the problems which will recur throughout

THE NUCLEOPHILICITY OF METAL COMPLEXES

3

this chapter. Close inspection of (3) shows that the coordinatively-unsaturated, square-planar iridium complex reacts with methyl iodide to give the coordinatively-saturated, octahedral complex in which both the methyl- and iodo-groups are separately bound to the metal. In accomplishing this reaction the formal Ir'') site has been oxidised to a formal Ir"") complex, but only because both the methyl- and iodo-groups are considered as anions when bound to the metal. It is because of this formalism that this type of reaction has become known as "oxidative addition", in which both the formal oxidation-state and coordination number have increased by two units. Ambiguities clearly arise from this formalism primarily because of the inability to define unambiguously oxidation-states, and because although some such reactions are oxidative (e.g. the reaction of CI, with [Pt""L,CI,] to give [Pt('")L,CI,]), others are not. as for instance in the reactions of hydrogen with tru~s-[IrCl(CO)(PPh,),1 as shown in (4), in which the classic reducing agent apparently oxidises the metal. Furthermore, in contrast to truns-[IrCl(CO)(PPh,),]

+ H,

-

[IrCI(CO)(H)2(PPh,),]

(4)

what is expected for an oxidising reaction, the addition of hydrogen to certain iridium complexes is favoured by electron-accepting ligands (Crabtree and Hlatky, 1980). Because of these conceptual problems a new notation for classifying these reactions has been proposed (Crabtree and Hlatky, 1980) in which the three-centre, oxidative additions of (3) and (4) are designated (3,2} reactions. The first figure in parentheses designates the number of centres involved in the reaction and the second figure gives the number of electrons transferred to the metal. The simple ligand addition shown in (5) is designated {2,2}.

+

[FeH(Ph2PCH2CH2PPh2),]+ py

-

[FeH(py)(Ph2PCH,CH,PPh,),]+

(5)

Although this system has the advantage of describing what is happening to the metal during the reaction its use is not essential provided one is conscious of the underlying assumptions in the formalism "oxidative addition". Thus in both reactions (3) and (4) the metal's d-orbital occupancy changes from d"-2 to d", but never fully becomes d" since there is (as a molecular orbital description reveals) a two-way flow of electron density. There is no requirement that the flow is balanced at any stage of the reaction (Saillard and Hoffmann, 1984). Both the positions of the coordinated carbon monoxide stretching frequencies in the infrared spectrum of the oxidative addition products of tram-[IrCI(CO)(PPh,),] (Vaska, 1968), and X-ray photoelectron spectroscopy of the adducts of [Ir(phen)(cod)]+ (phen = I , 10-phenanthroline; cod = 1,4-~ycloctadiene)(Louw et al., 1982) have been used to probe the metal-to-ligand and ligand-to-metal charge


4

transfer. In both studies the Me1 adduct has considerable metal-to-ligand charge transfer, and the “oxidation numbers” of the metal (2.58 and 2.48) are in good agreement with the picture described above. Having clarified what is meant by the term oxidative addition, it is clear that the description is purely stoichiometric and can cover a range of mechanistic pathways. The factors which increase the nucleophilicity of the site (i.e. a relatively low oxidation state and electron-releasing ligands) are the same factors which favour its behaviour as a reductant. As shown by the example in Scheme I , although a reaction may stoichiometrically resemble a nucleophilic pathway of the form of (l), this cannot be assumed and the simple second-order rate law (first-order in both reactants) cannot distinguish between the two mechanisms. Phf,

Pt-PPh,

PP+-Pt-PPh,

Pt-$P’

Ph,P\ Ph,P’

1

RX

Pt-X

+ ,

PPh,

slow

+

R’

1 PPh, R-Pt-X

Scheme 1

Obviously a discussion of the oxidative addition reactions which proceed by an electron-transfer mechanism is outside the scope of this review; for a thorough appraisal of this area the reviews by Halpern (1970), Lappert and Lednor ( I 976), Stille and Lau (1977), Deeming ( 1 972, 1974), Kochi (1 978), together with the recent article by Hill and Puddephatt (1985) and references therein, should be consulted. For the remainder of this chapter we will be concerned entirely with systems in which the nucleophilicity of the metal complex has been either unambiguously demonstrated, or at least strongly implicated. I t is important to be conscious of the spectrum of transition-state structures which could exist in the interaction between a metal centre and a molecule of R X as shown in Fig. 1.

5


R I

I

9 t.4

X I

I

‘M‘

M

FIG. I Spectrum of transition states for the reaction between a metal, M and alkyl halide. RX

At the extremes of this series the metal interacts with the carbon centre and the group X exclusively, whereas an intermediate form involves a three-centred transition state where equal interaction between the metal and both the carbon centre and X occurs. Strictly, only the transition states involving the interaction between the metal centre and the carbon centre are of relevance to the discussion herein, For this reason we have included briefly the mechanistic studies which are beginning to appear in the literature on the homogeneous activation of hydrocarbons. 3 Types of metal-based nucleophiles

Before discussing in detail the nucleophilicity of metal complexes towards organic molecules it is important to appreciate the depth of reactivity types that is being considered. Predominantly the organic substrates are alkyl halides and tosylates, acyl halides, aryl halides and ally1 halides together with a limited amount of information of the interactions with propargyl halides, and epoxides. It is no trivial task to classify the types of metal nucleophiles, but at present it is more sensible to categorise them in terms of their structure than their reactivity. In Table I the nucleophiles have been grouped according to their coordination number; this refers, however, to the “reactive” coordination number. Thus although the parent molecule is [Pd(PPh,),] or [Pd(PPh,),(C,H,)] the reactive species is the two coordinate [Pd(PPh,),]. In some of the complexes the n-bonded q 5-cyclopentadienyl-ligandis present, and is considered to occupy three coordination sites. For an introduction to the nomenclature and bonding of these x-type ligands the reader should consult Cotton and Wilkinson (1980). Two obvious points should be made about this table. The first is that derivatives of some of the complexes have not been included, such as the relatively trivial perturbation of using substituted cyclopentadienyl-ligands.Secondly, the inclusion of a particular complex in this table does not imply that aN its reactions with organic molecules proceed by a nucleophilic pathway, only that there is good evidence that it does behave as a nucleophile under certain circumstances.

SUNDUS HENDERSON AND RICHARD A

6

HENDERSON

TABLE 1 Types of metal nucleophiles Coordination number

3 4

Comment

Complex

[MRJ rrans-[MX(L) (PR,),] ~

[M(CO),I"

-

M = Pd or Pt, R = alkyl or aryl M = Sn, Ge or Pb M = Rh or Ir, L = CO or N,, X = halide M = Co, n = I ; M = Fe, n = 2; also protonated iron complex, [FeH(CO),J

-

M = Co or Rh, L = a range of neutral two-electron donor ligands M = Mn or Re n = 1, M = Fe or Ru, L = CO; n = 0, M = CO, Rh or Ir, L = CO, PPh, or q2-C2H4 M = Cr, Mo or W

5

The reactivity of these complexes will not be discussed individually, but rather the various features of the nucleophilic substitution reaction will be outlined and the complexes' behaviour will be compared and contrasted in this respect. 4

The scale of nucleophilicity

As will be seen in this section, metal complexes can be some of the strongest nucleophiles known, and they have been referred to as supernucleophiles. Furthermore, by subtle changes in the nature of the ligands, the nucleophilicity of the complexes can be varied. Thus although [RhCI(PPh,),] reacts readily with Me1 as shown in (6), substitution of one of the phosphine ligands by carbon monoxide to give tvans-[RhC1(CO)(PPh3),1 results in a molecule with a much diminished reactivity towards MeI. This difference in [RhCI(L)(PPh,),]

+ Mcl

-

[Rh(Me)CI(I)L(PPh,),]

( L = CO or PPh,)


7

reactivity is a consequence of the greater electron-withdrawing capability of carbon monoxide over the phosphine ligands, thus decreasing the electron density and hence the nucleophilicity of the metal centre (Collman and Roper, 1968 and references therein).

THE I N F L U E N C E O F L I G A N D S O N T H E N U C L E O P H I L I C I T Y

One of the most common ligands, particularly in this area of chemistry, is the monotertiary phosphine. This ligand is amenable to variation by simple perturbation of the alkyl- and aryl-substituents. Several studies have been reported in which the relative nucleophilicities of the metal complexes have been correlated, on a semi-quantitative basis, with the electron-releasing capability of the phosphine ligand. The rate of the reaction between Me1 and trans-[RhC1(C0)(PR3),] decreases with the nature of R in the order: n-C,H,, 2 n-C,H, > p-C,H,C,H, > p-C,H,C,H, > p-C6Hl3C6H4> C,H,. The analysis of the kinetic data for this reaction is complicated by phosphine dissociation from the complex and its subsequent quarternisation by Me1 (Franks rt ul.. 1981). The faster reactions with the alkyl substituents are a consequence of the greater electron-releasing capability of these groups. It has been proposed that as the number of carbon atoms increases beyond four, the basicity of the phosphine ligand is not markedly affected and so the trend observed for the first three phosphines reflects the steric influence of these groups. The influence of the p-substituents on the reaction rate is essentially that which would be expected on the basis of the electron-releasing effect of the substituent, except that the position of p-C,H,C,H, is slightly anomalous and this may represent a specific solvation effect. Similarly, studies on the reaction between Me1 and [Pt(PR,),] show an increase in rate in the order: R = p-C,H5C,H4 < p-MeC,H4 < n-C,,H,, < n-C,H, ,, consistent with the various electron-releasing capabilities of the p-alkyl groups. The faster reaction with the alkylphosphines again reflects the greater basicity of these phosphines. The increase in reaction rate with decreasing alkyl-chain length possibly has its origins in a steric effect, but more likely this is an electronic or specific solvation effect (Franks and Hartley, 1981). The kinetics of the reaction shown in Scheme 2 have been interpreted in terms of rate-limiting nucleophilic attack of [(qs-C,H,)M(CO)L] (M = CO, Rh or Ir, L = phosphine) on MeI. Rapid, iodide-induced intramolecular migration of the alkyl-group to the carbonyl-carbon atom yields the acylproduct (Hart-Davis and Graham, 1970). Variation of the phosphine ligand from PPh,, PPh,Me to PPhMe, leads to an increase in the reaction rate. The

a


Scheme 2

substituents on the phosphine are some three atoms away from the nucleophilic centre and thus the observed reactivity trend should reflect only electronic effects. However, despite P(C,H, 1)3 being one of the most basic phosphines used in these studies, the reaction with Me1 is one of the slowest, indicating that steric effects can play an important role, which even at three atoms removed from the reaction centre cannot be dissected from the electronic effects. This is further illustrated by comparison of the reactivities of [(qS-CsH,)Co(CO)L] (L = monotertiary phosphine) towards Me1 and the analogous nucleophilicities of the substituted pyridines as shown in Table 2. The nucleophilicities of the complexes [Co(L’H),L]- (L’H = substituted glyoximate ligand) have been studied as a function of varying the glyoximate ligand and the ligand L (Schrauzer and Deutsch, 1969). The nucleophilicities of these complexes towards Me1 depends on the energy of the 3di-orbita1, which can be influenced by the nature of the coligands. Studies on [Co(dmgH),L]- showed that the relative nucleophilicities decrease with L along the sequence: 2.6-lutidine > 2-picoline > C,H,,NH, > py > PhNH, > 4CNpy > Me,S > Ph,Sb > Bu,P > Ph,As > C,H,,NC > Ph,P. This series demonstrates the effect of both electron-donating substituents increasing the nucleophilicities and substituents with vacant orbitals decreasing the nucleophilicity. The decreasing nucleophilicity along the series corresponds to the decreasing donor, and increasing acceptor, strength of L. The “in plane” ligands (L’H) affect the nucleophilicity in the order shown below:

1;a;

\

OH

>

\

I

1 0-

0-

OH

> Phf

\

I

0-

9

Ph

>

l;i"

OH

\

\

I

0-

N 0-

TABLE2 Comparison of the relative nucleophilicities of substituted pyridines and [(q5-C,H,)Co(CO)L] ( L = phosphine) towards Mel"

A-w,

OC x-,,I

3.'

1 .0

,co\

PPh,

OC

I .0

6.7

/c*\ PPh,Me 6. I

P\ PPhMe,

OC

10.7

" Data of Hart-Davis and Graham (1970)

The strong inductive effect of the methyl-substituents renders dimethylglyoximate the most electron-releasing. The difference in reactivity between the cyclohexyl- and cycloheptyl-ligand has been attributed to the angular dependence of the orbital overiap between the nitrogen donor and the cobalt atoms, which is perturbed by the conformational preferences of the two rings. As would be expected, the formally analogous complexes containing the ligands [I] and [2] are weaker nucleophiles because of the decreased negative charge on the complex.

10


1

0..+y 0

The effect of changing either the axial ligand L or the “in-plane” ligand (L’H) is relatively small, but the latter has the greater effect. For instance [Co(dmgH),L] shows a change in rate of ca 50 in going from L = H,O to PR,. In contrast, the analogous species containing ligand [I] is somewhat anomalous exhibiting a factor of 500 for the same change in L (Toscano and Marzilli, 1984). In an extensive study of the reactions of Me1 with trans-[IrCI(CO)L,] (L = AsPh,, P(OPh),, PPh,, P(p-MeC,H,),, PMePh,, PEt, or PMe,Ph) the relative nucleophilicity of the complexes as a consequence of the phosphine was found to exhibit the same pattern of reactivity as already discussed. The arsine complex reacts only slightly faster than its phosphine analogue (Kubota et al., 1973). The rates for the reactions of Me1 with [IrYCI,] increase in the order Y = N, > PPh, > CO which reflects the electron-density at the metal; from N, to CO the n-accepting capabilities of these ligands progressively increase. Changing the anion in the complex trans-[IrX(CO)(PPh,),] augments the rate of reaction with Me1 in the order F % N, > CI > Br > NCO > I > NCS; however, the variation of reaction rate for the whole series only covers two orders of magnitude (Kubota et al., 1973). It has been pointed out that this order is that expected on the basis of the Hard Soft Acid Base (HSAB) postulate (Pearson, 1985). An interesting aspect of this series is the order of reactivity with respect to halide, or pseudohalide, which is the opposite of that found for the oxidative addition reactions of the same substrates with dioxygen, dihydrogen (Vaska, 1968; Chock and Halpern, 1966), benzenethiols (Gaylor and Senoff, 1972) and organic azides (Collman er a[., 1968). The more polarisable iodine atom renders the iridium metal the most basic, as determined by the protonation of trans-[IrX(CO) (PPhMe,),] with benzoic acid (Deeming and Shaw, 1971), and hence it would be expected that the iodo-complex would be the most nucleophilic. It has been proposed that the contrasting influence of the halido-groups on the reactivity of the complexes towards the various substrates is attributable to the different geometries (and hence different electronic requirements) of the transition states as shown in Fig. 2, where the


11

oxygen molecule is bound to the metal in a three-centre transition state, giving a pseudo-six-coordinate species, whereas with Me1 the transition state is only five-coordinate. X I

.

I

-1r-

-Ir/ \

/ \

FIG. 2 The different transition state structures tor the reactions between trnns[IrX(CO) (PPh,),] and oxygen or methyl halide

In a further study, the reactions of ~runs-[IrCl(CO)L,] ( L = PEtPh, or PEt,Ph) and trrms-[IrCl(CO){P(C,H,Z-p),),] (Z = CI, F, H, Me or MeO) with alkyl halides have been investigated (Ugo et a/.. 1972). The reactions with both methyl and benzyl halides were affected by the phosphines in the order: PEtPh, > PEt,Ph > PPh,. The anomalous position of PEt,Ph has been ascribed to steric effects, and this is consistent with the lower entropies of activation for reaction (7) when L = ethyl-substituted phosphines. frcrns-[IrCI(CO)L,] + RX

A,

[IrCI(X)(R)(CO)L,]

I

I

-0.2

-03

00

01

(7)

I

0.2

OP

FIG. 3

Hammctt plots for the reactions of truns-[lrCI(CO)(PAr3),] with PhCH,CI

( A )and Me1 ( A )

12


The p-substituted arylphosphine ligands represent an ideal system to study the electronic effects of reaction (7) since they are uncomplicated by steric effects. Marked differences in the reaction rate with both Me1 and PhCH,CI are observed with a change of the p-substituent. A graph of log,, k , where k , is the second-order rate constant for reaction (7) against the Hammett oP constant yields two lines as shown in Fig. 3, for the two alkyl halides. The reaction constants (p) are -6.4 (MeI) and - 2 . 6 (Ph,CH,CI). The large negative value with Me1 is indicative of a positive charge formed at the iridium atom in the transition state. The less negative value for PhCH,CI demonstrates that the binding of the benzyl group to the metal in the transition state is weaker than is the binding of the methyl group. THE N U C L E O P H l L l C l T Y O F METAL COMPLEXES

Relatively little systematic work on the nucleophilicities of metal complexes has been performed. Of particular interest would be the definition of a scale of nucleophilicity based on isoelectronic metals with the same coligands. The variety of different types of nucleophiles and the dichotomy of electron transfer vs nucleophilic mechanisms precludes any extensive comparison. However, it is possible to compare the relative nucleophilicities within a given group. The relative rates of reaction between Me1 and [(q5-C,H,) M(CO)(PPh,)] (M = Co. Rh or Ir) are in the order: Co(l.O), Rh(1.4), Ir(8.0) (Hart-Davis and Graham, 1970). whereas for the reaction between pentafluoropyridine and [M(CO),(PPh,),J the order is : Co( 1 .O), Rh(857), Ir(357) (Booth et al., 1975). Although both studies demonstrate the greater nucleophilicity of the heavier members of the triad, it is clear that 'the relative reactivities of the members can vary. Such behaviour is also manifest in the acid strengths of transition metal hydrides (Pearson, 1985; Shriver, 1970; Jordan and Norton, 1982 and references therein; Ziegler, 1985). Comparison of [M(dmgH),OH]- (M = Co or Rh) in its reaction with Me1 has a Pearson nucleophilicity of Co(14.3) and Rh(13.7) (Schrauzer et at., 1968; Weber and Schrauzer, 1970). In an early study (Dessy et al., 1966), reduction of homobimetallic complexes at a mercury electrode was used to generate the corresponding mononuclear anions in situ as exemplified in (8). Although the reaction

13


products were not isolated, studies of the reaction between a wide variety of nucleophiles and several alkyl halides were determined semi-quantitatively. The span of reactivity observed covers more than 12 orders of magnitude, with the nucleophilicity of PhS- lying in the centre of the scale. A linear relationship between log,,, k, and - E i o x (measured at a Pt electrode) is observed as defined by (9). The parallelism between nucleophilicity (and basicity) and the oxidation potential of a complex is a consequence of a nucleophilic displacement (or protonation) at the metal formally resulting in an oxidation of the metal in the transition state (or product). log,, k ,

-

=

6.6EioX-3.0

(9)

ee Fe

[Pt(PEt,),l

I

y,

oc' 'co

LU

ee-

w oc k c o / \

ho

[ColdmgHipyf [RelCOISS

/ \

6

IPt lPPhJ31

[Mn(C0151-

OCLCO c,

1Rh(CN i1-,

6

el

oc':r'co C

0

[ coca), 1-

PhsPPh,

FIG.4 Pictorial representation of the nucleophilicities of various transition metal complexes. The scale is log,, (k,/koTs),where k, is the second order rate constant for the reaction of the nucleophile with MeI. and k,,, is the corresponding value for MeOTs

14


HENDERSON

More recently an extensive study on the nucleophilicities of several transition metal complexes has been reported (Pearson and Figdore, 1980). Experimental difficulties over the range of nucleophiles studied precluded the use of a single solvent. Because of the range of solvents employed, and the various influences on the reaction rates with variously charged nucleophiles (Section 3, i t is not easy to establish an absolute scale of nucleophilicity. However the accumulation of data presented by Pearson and Figdore together with other studies has led to the nucleophilicity scale shown in Fig. 4 (Collman ef ul., 1977; Hart-Davis and Graham, 1970; Douek and Wilkinson. 1969; Pearson an d Gregory, 1976; Collman and MacLaury, 1974 and Dessy ef d.. 1966). The collected data give the order of transition metal nucleophiles as follows: Ni(0) 9 Pd(0) % Pt(O), (for the species [M(PR,),]) and Fe(0) > Ru(O), (for the species [(q5-C,H,)M(CO),]-), Re( - 1) > Mn( - I), (for the species [M(CO),]-), Ir(l) > Rh(l) > Co(l), (for the species [(~5-C,H,)M(CO)(PPh,)]) and W(0) > Mo(0) > Cr(O), (for the species [(q’-C,H,)M(CO),] -). Clearly there is no simple criterion for establishing the reactivity of a given “complex-type” even within the same group. The reluctance of alkyl sulphonates to react by a free-radical pathway has led to a proposed kinetic distinction between nucleophilic and electrontransfer pathways (Pearson and Figdore, 1980). Failure to react with methyl tosylate (MeOTs), but rapid reaction with Me1 could be a criterion of the electron-transfer pathway. For the reactions of a large range of metal complexes there exists a linear correlation between log,, kZMeoTs and log,, kZMe’indicative of an S,2 mechanism common to all the reactions. However, there are notable exceptions to this graph, such as Li,[PtMe,], which does not lie on the line despite strong indications of its nucleophilicity. Further corroboration of a common mechanism for all these complexes is the observed isokinetic plot. However the closeness of the isokinetic temperature to room temperature ( T = 282 f 22 K) and the variety of solvents used in these studies make any such conclusions from these correlations highly dubious. Indeed [Co(CN),I3- also lies on this isokinetic plot, but is known to react with Me1 by halogen-atom abstraction (Chock and Halpern, 1969). As indicated throughout this section, the nucleophilicity of the complexes is critically dependent upon the nature of the solvent, and this aspect will be discussed in more detail in the next section. 5 The influence of the solvent on the nucleophilicity

Surprisingly, relatively little systematic work has been performed aimed specifically at understanding the influence that the solvent has on the nucleophilicity of the complexes. An early study (Schrauzer and Deutsch,

THE NUC LEO PH I LI C ITY 0 F M ETAL CO M PLEX ES

15

1969) on the reactivity of vitamin B,, with MeCI, and of [Co(dmgH),(PBu,)]- with EtBr, concluded there was no large solvent effect on these reactions in methanol, ethanol, propan-1-01 and H,O, together with some mixtures. However, the range of solvent types covered is clearly very narrow, and subsequent, more extensive, studies have demonstrated that the reactions of many transition-metal nucleophiles are facilitated by polar solvents. The rate of reaction of [(q5-C5H,)M(CO)(PPh3)](M = Co or Rh) with Me1 is very dependent on the solvent as shown in Table 3 (Hart-Davis and Graham, 1970). Studies on the reaction of [PtPh,(bipy)] (bipy = 2,2'-bipyridine) with Me1 or PhCH,Br in a range of solvents (benzene, ethyl acetate, dichloromethane, acetone or nitromethane) showed, as with [(q5-C5H5)Co(CO)(PPh3)]and trans-[IrCI(CO)(PPh,),], that a factor of 10-15 in the reaction rate was observed in going from the least to the most polar. This implies a common nucleophilic reaction for all the complexes (Jawad and Puddephatt, 1976). TABLE3 The influence of the solvent on the rate of [(q'-C,H,)M(CO) (PPh,)] (M = Co or Rh) and Me1

THF"

Solvent

Me,CO

the

reaction

CH,CI,

between

CH,CN ~~

Dielcctric constant M = CO. k,,, M = Rh, k,,,

"THF

=

7.4 0.2

20.7 0.9

9.08 I .o 49.3

cu

36.0 2.5 1 .o

tetrahydrofuran

Early work comparing the reactivity of trans-[IrC1(CO)(PPh3),1towards the substrates 0,, H, and Me1 (Chock and Halpern, 1966) demonstrated that the greater sensitivity of the reaction with Me1 to a change in solvent was a consequence of a less negative AS', compensated only in part by an increase in A H * . Thus, in the strongly ionising solvent N,N-dimethylformamide (DMF), the reaction with Me1 is seventeen times faster than that in benzene at 30°C. This is similar to the characteristics of the Menschutkin reaction, and thus the reaction of trans-[IrCI(CO)(PPh,),] with Me1 was concluded to proceed via the five-coordinate transition state shown in Fig. 1. Further work on the reactions of rrans-[IrC1(CO)(PPh3),] with Me1 in a variety of solvents corroborated these conclusions (Ugo et al., 1972). The qualitative observations that the reactions of trans-[IrC1(CO)(PMe3),] with CH,CHBrCO,Et and PhCHFCHBrC0,Et are faster in the more polar

16


solvents such as N-methylpyrrolidone (NMP) and D M F than in benzene raises doubts on the use of such a probe as a mechanistic criterion, since these reactions proceed by an electron-transfer mechanism (Labinger and Osborn, 1980) (see Section 6). The problems in defining a nucleophilicity scale using a variety of solvents have been outlined (Pearson and Figdore, 1980). The influence of the solvent on the nucleophilicity of neutral complexes such as trans-[IrCI(CO)(PPh,),1 with Me1 is relatively simple. For instance, the reaction in tetrahydrofuran (THF) is three times faster than it is in benzene (Ugo et al., 1972). However, when the nucleophile is anionic, then comparison of the reactivities in two different solvents is not simple. Electrostatic arguments would indicate that a free anion would be more reactive in a medium of lower dielectric constant, but extensive ion-pairing with counter cations greatly augments the reactivity. Several studies on the nature of some transition-metal nucleophiles in solution have been reported, and these give important insight into the reactivities of these species. Detailed studies of the solution infrared spectrum (carbonyl stretching region) have established the sites and extent of alkali-cation interaction with [(q'-C,H,)Fe(CO),]- (Pannell and Jackson, 1976; Nitay and Rosenblum, 1977); [Mn(CO),]- (Darensbourg et af., 1976; Pribula and Brown, 1974); [Co(CO),]- (Edgell er al., 1978; Edgell and Chanjamsri, 1980); yV(CO),L](L = PPh,, P(OPh),, P(Bu"), and CNMe) (Darensbourg and Hanckel, 1981. 1982) and [(q5-C,H,)M(CO),]- (M = Cr, Mo or W) (Darensbourg et d..1982). In all cases the cations are closely associated with the carbonyl oxygen atoms. Simplistically it might be considered that ion-pairs would always be poorer nucleophiles than the "bare" anions. However, quite the opposite has been observed in the reactions of epoxides with [(q5-C5H5)Fe(C0)J (Nitay and Rosenblum, 1977). the reactions of benzyl halides with [Co(CO),]- (Moro et ul., 1980) and [Mn(CO),L]- (L = CO, PR, or P(OR),) (Darensbourg et ul., 1976) and [(q5-C5H,)Mo(CO),]- (Darensbourg ef a/., 1980). These results have been interpreted in terms of a highly-ordered transition state in which the cation assists carbon-halogen cleavage as shown in Fig. 5.

FIG. 5 Highly ordered transition state for the reaction between a metal carbonyl complex and alkyl halides in the presence of alkali metal cations


17

The reaction of [ ( I ~ ~ - C ~ H ~ ) M O ( Cwith O ) ~PhCH,CI ]in T H F is more rapid in the presence of sodium ion, and the addition of hexamethylphosphoramide (HMPA) (which coordinates to the sodium) inhibits the reaction. However, the analogous reactions of Bu"1 show the opposite effect. These effects are a consequence of the local environment of the nucleophile, rather than a bulk medium effect. Reaction (10) between [(q5-C5H5)W(C0)3]-and CH3C= C(CH,),I is faster, however, when the tungsten complex is present as the (Bu,N+) salt rather than as the Na' salt, presumably because of the greater nucleophilicity of the complex in the presence of the more chargedispersed cation (Watson and Bergman, 1979).

It is possible to investigate much purer solvent effects using the large bis(tripheny1phosphine)iminium cation (PPN +), which, although associated with [(q5-C5H5)Mo(C0),]- in THF, does not electronically perturb the carbonyl groups. The second-order rate constants for the reactions between [ ( T ~ ~ - C ~ H ~ ) M O ( and C O )Bu"I ~ ] - or PhCH,CI in acetonitrile and T H F are shown in Table 4. These values give further credence to the proposed SN2 mechanism. The charge separation present in a transition state of SN2 character would be less stabilised in a polar solvent than the ground state of an anionic nucleophile and neutral alkyl halide, giving rise to a correspondingly higher activation energy. The data observed for the reactions with PhCH,CI indicate a greater degree of charge separation in the transition state with this substrate. TABLE 4

Kinetic data for the reactions of (PPN) [(q5-C,H5)Mo(CO),]with alkyl halides"

' '

1 04k ./dm3mol - s -

Solvent

Bu"I

PhCH,CI

MeCN TH F

1.64 26.0

1.60 0.63

a

Darensbourg et d., 1982

A detailed study on the influence of the solvent on the nucleophilic reactivity of Na,[Fe(CO),] towards alkyl halides and sulphonates, and the dramatic effect that ion-pairing has on the reactivity, have been reported (Collman et al., 1977). The reactions studied are typified by ( 1 I).


18

0

In NMP or T H F the values of the two successive ion-pair dissociation constants of [Fe(CO),I2- shown in (12) were estimated from a combination of conductometric titrations in the presence and absence of the cryptand, Na,[Fe(CO),I

A,,

ZIIII?

Na'

+

Na[Fe(CO),]-

K,,

Z L IZ

2Na'

+ [Fe(C0),l2-

(12)

Kryptofix 2,2,2. The corresponding values and limits estimated from this study were: in NMP, KID = 0 . 2 8 , 1 x l o W 3> K , , > 1 x m~l.dm-~; K,, @ 5 x mol.dmP3. in THF, K , , - 5 x The mechanism of the reaction with various alkyl halides in NMP is believed to involve nucleophilic attack by both the solvent-separated ion-pairs Na,[Fe(CO),] ( k , ) and [Na(Fe(CO),]- (k',) (where k', = 0.18 dm3 m o l - ' s - ' and k , < f k',) as evidenced by the non-linear dependence of the observed rate constant on the concentration of complex. In T H F the dissociation of the ions is less extensive, and the rate of reaction with alkyl halides is correspondingly slower. However, the addition of polar solvents or crown ethers increases the rate of the reaction, and this has been ascribed to the kinetically more reactive, more dissociated species. 6

Stereochemical changes at a saturated carbon centre

In general, the most convincing evidence that the reaction of an alkyl halide or tosylate with a metal complex adopts a nucleophilic displacement (S,2) mechanism is the demonstration that inversion of the configuration of the saturated carbon centre has occurred as a consequence of this process. However, as will be seen, inversion of the configuration of the carbon atom is not an essential consequence in the reactions of coordinatively-unsaturated. transition-metal nucleophiles. REACTIONS OF COORDINATIVELY-SATURATED COMPLEXES

In practice, it is not easy to establish the stereochemical consequence of the reaction between a chiral carbon centre and a metal complex because of the lack of crystallographic information establishing the absolute configurations of metal-alkyl complexes. In order to determine the stereochemistry of the addition, one strategy is subsequently to cleave the metalbcarbon bond with a reagent so as to regenerate the original organic compound. Provided


19

that the stereochemistry of the cleavage reaction is known, the stereochemistry of the nucleophilic displacement reaction can be established. As an example, the sequence of reactions shown in Scheme 3 demonstrates the inversion of configuration of the nucleophilic displacement between [(q5-CsH,)Fe(CO),]- and MeCHBrC,H,, but only because it is implicitly assumed that the cleavage of the iron-carbon bond occurs without affecting the configuration of the chiral carbon centre (Johnson and Pearson, 1970).

Me

PPh,

retention

Scheme 3

Examples are known, however, in which halogen cleavage of metal-alkyl bonds occurs with retention (Whitesides and Boschetto, 1969; Calderazzo and Noack, 1966; Johnson, 1970; Slack and Baird, 1974, 1976) and others with inversion (Whitesides and Boschetto, 1971; Jensen and Davis, 1971; Jensen et a/., 1971; Anderson et al., 1972). Without firm evidence about the stereochemistry of the cleavage of metal-alkyl bonds, little can be said about the stereochemistry of the reaction between metal nucleophiles with alkyl halides. A further complication that can occur in these stereochemical studies was demonstrated in the reaction between [Mn(CO),]- and optically active MeCHBrCO,Et, where subsequent bromine cleavage was used to establish the stereochemistry of the reaction. The stereospecificity observed in this


20

7

reaction is low (ca 60%) and this is a consequence of racemisation of the alkyl-complex in the presence of an excess of [Mn(CO)J as shown in (13) (Johnson and Pearson, 1970).

+ [Mn(CO),I -

(13)

The elegant work of Whitesides and coworkers represents an unambiguous means of determining the stereochemistry of the reaction between nucleophiles and saturated carbon centres. The work centres around the stereospecific preparations of erythro- and threo-3,3-dimethylbutan1-01- 1 ,2-d,[ButCH(D)CH(D)0H].The two isomers are readily distinguished by their characteristic vicinal coupling constants {J(erythro) = 9.5 Hz; J(threo) = 5.8 Hz} (Whitesides and Boschetto, 1969; Block et al., 1974). In the original studies, conversion of the erythro-alcohol to the bromobenzenesulphonate (with retention of configuration) and subsequent reaction with [ ( T ~ ~ - C , H ~ ) F ~ ( C produced O ) ~ ] - a product whose 'H nmr spectrum consists of a single AB quartet 61.38, J = 4.4 Hz. Independent analysis of the vicinal coupling constants for both the complexes containing the threo- ( J = 4.6 Hz) and erythro-ligand ( J = 13.1 Hz) showed that the reaction between erythro-Bu'CH(D)CH(D)SO,C,H,Br and [(q5-CsHs)Fe(CO),]- occurs with inversion of configuration at C-l as shown in Scheme 4.

D

Fe

I\ oc co

D

D

H

OBs

threo -complex

erythro Scheme 4

Subsequent studies, occasionally using the trifluoromethane sulphonate derivative. have extended this work to less potent nucleophiles such as


21

[(qS-C5H5)Mo(C0)J-, [Co(dmgH),py] - , C,H5Se- and Me,Sn- (Block and Whitesides, 1974). In all the studies, except those with the tin compound, inversion of configuration at C-l was found with a stereoselectivity of greater than 95%. However, with Me,,%- only 80% inversion was observed, a point to which we shall return later. Another area where 'H nmr spectroscopy can be used to determine the stereochemistry at the carbon atom is demonstrated in the reactions of substituted cyclohexanes with [Co(dmgH),py]-, as shown in the selective examples (14) and ( 1 5). The stereochemistry of the displacement (inversion of the configuration of the carbon centre) and the strong preference for the metal residue to occupy the less congested equatorial site, defines the orientation of the substituent Y at the 4-position.

a Co(dmgH),pyl-

+Y

x

-

1

[ y ~ o ( d m ~ H ) 2 P Y x- ( ]

\

(X

=

Y

=

Br or X

=

OTs, Y

=

+

OH)

Demonstration of the configurations of the products shown in (14) and (15) comes from comparison of the 'H nmr spectra of the products with those of the formally analogous t-butyl-substituted cyclohexyl bromides, where the t-butyl-group, like the metal residue, is forced into an equatorial position (Jensen et a[., 1970). Alternatively assignments can be made on the basis of the magnitude of the coupling constants: J(ax-ax) = 8-14Hz > J(eq+q) or J(ax-eq) = 1.7 Hz. However, the assignments based on these coupling constants are less compelling, amounting to a distinction based on the resolution of a signal into a doublet or the observation of a broad singlet. Finally the reaction of a-bromocamphor with isotopically-labelled [2HFe(CO),]- gives rise to camphor-3-exo-d as shown in ( I 6). This result

&;:

.

& L O

I*HFc(C'O)J

Br

(16)

H

has been interpreted (Alper, 1975) in terms of an S,2 attack at the carbon atom, with inversion of configuration. As discussed before, however, in the absence of any evidence concerning the sterkochemistry of the subsequent cleavage, this conclusion must remain tentative.

22


REACTIONS O F C O O R D I N A T I V E L Y - U N S A T U R A T E D COMPLEXES

For the reactions of coordinatively-saturated nucleophiles with alkyl halides, the overwhelming evidence is that they proceed with inversion of configuration at the carbon atom as expected for an SN2-typereaction. With the square-planar, coordinatively-unsaturated reactants such as trans[IrCI(CO)(PPh,),], however, the possibility exists that either inversion or retention of configuration at carbon can occur, depending upon whether the transition state is five-coordinate or six-coordinate with a “side-bonded’’ approach of the alkyl halide to the metal centre (Fig. 1). Indeed this area has been one of some controversy over the years. In 1970 two conflicting reports on the stereochemistry of the addition of chiral alkyl halides to square-planar iridium(1) complexes appeared. In one report it was claimed that the reaction of trans-[IrCI(CO)(PPh,Me),] with optically active CH,CHBrCO,Et occurred with retention of configuration as shown in Scheme 5 (Pearson and Muir, 1970). This result is consistent with a “six-coordinate intermediate”, as is the lack of incorporation of any free halide into the product (Section 8). However, the conclusions should again be treated with care since the study employs the cleavage of the iridiumxarbon bond by halogen, and without knowing the stereochemistry of this reaction little can be said about the stereochemistry of the displacement. MePh,P

\

CH,CHBrCO,Et

’

CI /Ir\P0\Me

[,I&’

~

-6 0’

IIr(Cl)(Br)lCHMeCO,Et)(CO)(PPh,Me),I

[QK--20’

I

Br,.

~

78’C. THF

CH,CHBrCO,Et I:[. Scheme 5

-4.0’

THE NUCLEOPHlLlClTY OF METAL COMPLEXES

23

In contrast, the reaction of trans-[IrCI(CO)(PMe,),] with trans-2-fluorocyclohexyl bromide was claimed to involve inversion of the configuration at the carbon atom as shown in (17) on the basis of I9F nmr spectroscopy of the product (Labinger et al., 1970). Br

+ [IrCI(CO)(PMe,),]

-

[e

IrCl(Br)(CO)(PMe,),

(17)

F

A subsequent study of reaction (17) claimed that under the conditions reported by Labinger et al. no reaction occurred (Jensen and Knickel, 1971) and that the infrared spectrum of the reaction mixture was sensitive to traces of radical initiators (e.g. dioxygen). However, more detailed studies of the reactions between a variety of alkyl halides and trans-[IrCI(CO)(PR,),] has clarified this confused literature. In these studies the alkyl group contains a fluorine atom which can be used as the nmr probe (Bradley et al., 1972; Labinger et al., 1973; Labinger and Osborn, 1980). The contentious studies with trans-2-bromofluorocyclohexane were repeated with trans-[IrCI(CO)(PMe,),[ and the products isolated and purified fied. The correspondingly much improved 19F nmr spectrum (compared to the previous study) now demonstrated that two products were formed. Furthermore, the reaction of cis-2-bromofluorocyclohexane with trans[IrCI(CO)(PMe,),] yielded an identical 19Fspectrum clearly indicating that complete loss of stereochemistry at the carbon atom had occurred. The two most likely products are [3] and [4].

[31

[41

Loss of stereochemical integrity has also been observed in the reactions of trans-[IrCl(CO)(PMe,),] with 1 -bromo-2-fluoro-2-phenylethane- 1 -d,. The reaction of the ( R R , S q - and (RS,SR)-diastereoisomers produces the same product mixture, demonstrating the complete racemisation at the carbon centre. Studies on (RR,SS)- and (RS,SR)-ethyl-3-bromo-3-fluoro-3-phenylpropionate with rrans-[IrCI(CO)(PMe,),] also demonstrated that the same


24

products were formed from both isomers in a non-stereospecific process. This result brings into question the proposed retention of configuration observed in the reactions of MeCHBrC0,Et and fr~ns-[IrCl(CO)(PPh,)~] (Pearson and Muir, 1970). Using a much improved preparation of ethyl-(R)-( )-a-bromopropionate, the reaction with trans-[IrCl(CO)(PMe,),] yielded a product in which complete racemisation occurred at the carbon centre. These reactions were all repeated with a series of complexes of formula, trans-[IrCI(CO)L,] (L = PMe,Ph, PPh,Me or AsMe,Ph) and in all cases complete loss of stereochemical integrity at the chiral centre was observed. One possible means of racemisation of the alkyl complex containing an a-hydrogen, is rapid proton exchange as shown in (18). However, this

+

possibility was eliminated since the reaction of ethyl (R)-( +)-a-bromoproin the presence of MeO'H shows no pionate with tr~ns-[IrCl(CO)(PMe,)~] incorporation of deuterium into the product. Furthermore the isolation of excludes one diastereoisomer of [1r(PhCHFCHC0,Et>C1(Br)(C0)(PMe3),] the possibility of rapid epimerisation. Changing the solvent from benzene to the more polar NMP and D M F does not affect these stereochemical results. These stereochemical results, together with some further mechanistic studies on the same reactions (Labinger et al., 1980) have led to the conclusion that the reactions of saturated alkyl halides (except methyl derivatives), vinyl and aryl halides and a-haloesters with trans-[IrCl(CO)(PR3)J proceed by a radical chain pathway, whereas methyl, benzyl and ally1 halides and a-haloethers probably react by a nucleophilic displacement process. One final type of coordinatively-unsaturated complex where stereochemical studies have demonstrated the nucleophilic character of the metal towards alkyl halides is [Pd(PR,),]. These reactions contrast with those of the analogous platinum complexes where the participation of free radicals has been demonstrated (Stille and Lau, 1977 and references therein). The reaction of cis-3-acetoxy-5-carbomethoxycyclohexenewith [Pd(PPh3)3], and then cleavage of the product with methylphenylsulphonylacetate gives the cis-cyclohexene product as shown in (19) (Trost and Strege, 1977). The formation of this isomer is a consequence of two inversions, since the reaction of the intermediate n-bonded palladium


Ph,P

25

PPh,

complex with the acetate has been shown to occur with inversion of configuration, thus defining the SN2-type attack of [Pd(PPh,),] on the cyclohexene derivative. The reactions of some optically active benzyl halides have provided further evidence for the nucleophilicities of the [Pd(PR,),] species (Lau et al., 1974, 1976; Wong et al., 1974; Stille and Lau, 1976). The salient features of these studies are summarised in Scheme 6. The stereochemistry of the addition is established by reaction of the palladium-alkyl complex with carbon monoxide (this “insertion” is known to take place by an intramolecular migration process, with retention of configuration in the migrating alkyl group), and subsequent formation of an ester from this acyl complex. Ph

\

TPh,

Ph 0 PPh, 1 II I -C- C - Pd - X I R’cH PPh,

C-yd-X R-i H PPh, Pd(PPh,),

YOH

/

HF .

Ph- C,

X

/

\

IPd(PPhJJCOl1

Ph \ C;R, H

$0

- F;d - X PPh, Scheme 6

These subsequent reactions occur at a position remote from the chiral centre. In this way the additions were shown to proceed quantitatively (or essentially so) with inversion of the configuration of the carbon centre.

26


However, in the analogous reactions of [Pd(PEt,),] and PhCHDX (X = C1 or Br), some loss of stereochemical integrity (ca 30%) is observed (Becker and Stille, 1978). This has been ascribed to rapid nucleophilic exchange after the formation of [Pd(CHDPh)Cl(PEt,),] which inverts the configuration of the alkyl-group. When the substrate is PhCHDBr only 19% net inversion is observed and the isotopically labelled bibenzyl (PhCHDCHDPh) is an isolable product. This result has been attributed to a competitive, one-electron transfer reaction leading to the radical pair shown in [5]. Ph

151

A less clear-cut demonstration of the nucleophilicity of [Pt(PPh,),] comes from its reaction with optically active 8-(a-bromoethyl)quinoline, as shown in (20). The configuration of the product was deduced by Brewster’s rules (Sokolov, 1976). However, prior coordination of the nitrogen atom to the metal could predefine the pathway.

+ [Pt(PPh,),]

-

+ 2PPh,

‘

H---C-Pt-Br

Me

(20)

I PPh,

Although it has been deduced from several studies that the reaction between [Pd(PR3),] and ally1 acetates occur with inversion of the configuration of the carbon atom (Trost and Verhoeven, 1978, 1980), it is only recently that direct proof has been obtained in the reaction of [Pd(PPh,)( Ph,PCH,CH,PPh,)] with (S)-(E)-3-acetoxy-1-phenyl- 1-butene as shown in (21). The gross stereochemistry of the product was established by ‘H nmr spectroscopy, and measurement of its optical rotation, and comparison with that of the optically pure isomers prepared by an alternative route, demonstrated that the reaction occurred with 8 1 Oh, inversion of the carbon’s configuration (Hayashi et al., 1983).

Tph +

[Pd(PPh,)(Ph,PCH,CH,PPh,)]

OAc

-

PPh,

t

+ AcO- +

Pd

phzp


REACTIONS OF

27

Me,Sn-

The reactivity of Me,Sn- has been reserved for a separate section, since with this species particularly the electron-transfer and nucleophilic pathways are energetically very similar. In this section we will not review in detail the somewhat controversial literature concerning the electron-transfer mechanisms; for this the reader should consult Kitching er al. (1978); and references therein. Instead, only those studies relevant to the compound’s nucleophilicity will be discussed. As indicated earlier, the 80% inversion of configuration at the carbon atom observed in the reactions of Bu‘CHDCHDOTs with Me,Sn- (Block and Whitesides, 1974) should only be considered as a minimum value. However the stereochemistry of the reactions of alkyl compounds with Me,Sn- has generally been thought to be a straightforward indicator of the mechanism; the nucleophilic pathway gives a product of inverted carbon configuration whereas the electron-transfer pathway yields a racemic product. The demonstration of free-radical reactions which proceed with inversion of the configuration at carbon (Kuivila and Alnajjar, 1982; Ashby and DePriest, 1982) must cast some doubt on this stereochemical distinction between the two pathways. The reaction of Me&- with 2-butyl bromide has been shown to proceed with inversion of the carbon’s configuration (Jensen and Davis, 1971), but many subsequent stereochemical studies have shown more complicated behaviour. The reactions of Me,Sn- with cis-4-t-butylcyclohexyl bromide and with trans-4-t-butylcyclohexyl tosylate have been shown by H nmr spectroscopy to give the same product, and on the basis of independent evidence this product has been assigned the cis-configuration as shown in Scheme 7. Thus the reaction of the tosylate occurs by a nucleophilic displacement reaction with inversion of configuration at the carbon centre, and the bromide reaction apparently proceeds with retention (Koermer et al., 1972).

retention

1 Me,%-

/ inversion

Scheme 7

SnMe,

28


Furthermore, the reaction of Me,Sn - with I -bromo- 1 -methyl-2,2-diphenylcyclopropane is said to occur with retention of configuration at the carbon atom, based on the net retention upon HCI quench of the stannylcomplex (Sisido ef al., 1967) as shown in (22). The reactions occurring with retention presumably proceed via a four-centre transition state.

Ph

'nph Mh

HCI

~

hie

Ph

Ph

The stereochemistry of the reaction between 7-bromonorbornene and Me&- is strongly dependent upon the solvent (Kuivila er al., 1972). In particular it is observed that, as the coordinating capability of the solvent towards the cation (associated with Me,Sn-) increases, the proportion of inversion at the carbon centre increases. In a detailed study of the reactions of Me&- with cis- or trans-4-t-butylcyclohexyl bromides in THF, mixtures of products were obtained with complete loss of stereochemical integrity at the carbon atom, thus implicating radical reactions (San Filippo Jr el al., 1978). Similar studies of Me,M(M = Ge or Sn) with 4-alkylcyclohexyl bromides and the corresponding tosylates showed that the latter react with inversion of configuration, consistent with an S,2 mechanism, whereas the bromides react via a non-stereospecific process for which several possible pathways have been discussed (Kitching et al., 1978). The reactions of cyclopropylmethyl bromide proceed via free, non-caged, radicals. The stereochemistry of the reaction is also sensitive to the countercation of Me&-. Thus the lithium salt at -70°C gives a higher proportion of the product with inverted configuration than the sodium salt. This indicates that at low temperatures there is a nucleophilic pathway operating. The low-energy radical pathway for the reactions of Me,Sn- is consistent with the observations (Koermer er al., 1972) that olefins are sometimes observed as minor products of the reactions shown in Scheme 7. More recently, studies of the reactions of (R)-2-C8H,,X (X = C1, Br or OTs) with Me,SnLi have shown that the stereoselective inversion of configuration decreases along the series OTs > CI > Br and, furthermore, the temperature and mode of addition affects the selectivity (San Filippo Jr and Silbermann, 1981). There has also been criticism of earlier work in which trapping agents were used to detect radical reactions; it has been claimed that the trapping agents perturb the reaction and give rise to radical mechanisms.


29

REACTIONS OF BINUCLEAR COMPLEXES

An interesting, and to date unique, study of the stereochemical consequences of the reaction between an alkyl halide and a binuclear nucleophile comes from the study of the reactions of [{(q5-C5H5)Co}2(p-C0)2]-. This species reacts with methyl iodide to give the binuclear product shown in (23).

However, it is difficult to investigate the stereochemistry of this simple reaction because reactions with alkyl halides involving alkyl groups larger than ethyl d o not lead to stable products.

The stereochemistry of the metallacycle formation was determined as shown in (24). The alkylation of the dicobalt anion with a,y-diiodides occurs with a complicated stereochemistry as evidenced by the studies with mesoand (f)-2,4-dimethylpentane. The mechanism of the reaction is shown in Scheme 8. The initial alkylation occurs with complete loss of stereochemical

FI

-

/c\

cpco=cocp

‘c’

+

yy H

6

I

c7 cpco=cocp

k I

+

‘c’

H I

(A)

d

cpco-

cocp

I ’ \\

&’ $ 0

0

Scheme 8

c;co-c~cp

1’1 $ $ 0

0

30


HENDERSON

integrity as a consequence of an electron-transfer process. Subsequent reduction of the monoalkylated product is followed by ring-closure with predominant inversion of configuration, presumably because of an SN2 mechanism (Yang and Bergman, 1983). 7 Stereochemical changes at an unsaturated carbon centre REACTIONS OF COORDINATIVELY-SATURATED COMPLEXES

A detailed study of the reactions between alkynyl or vinyl halides and

[Co(dmgH),py] - has demonstrated a complex pattern of behaviour (Cooksey et al., 1972). At low concentrations of nucleophile over short reaction times, the reaction with prop-2-ynyl halides yields the prop-2-ynylcobaloxime, presumably by a nucleophilic displacement pathway. However, at longer reaction times in the presence of an excess of cobaloxime, allenylcobaloxime is observed. This is formed as a consequence of an SN2' displacement of one cobaloxime by another attacking the initially formed prop-2-ynylcobaloxime as shown in Scheme 9. The reactions of substituted

-Br-

(CO);

+

BtCH, C

CH

(tor

(Co),-CH,C= CH 2 CH2=C=CH-(CoIZ

+

(CO);

s, 2'

SN2

Scheme 9

prop-2-ynyl halides results only in the substituted allenyl products as shown in (25). This is a consequence of the hindered access to the saturated carbon centre making the direct S,2' mechanism the one of lowest energy. Similar conclusions have been reached for the reactions of ally1 halides with the cobalt nucleophile. Thus the reaction of either a- or y-methylallyl halides with [Co(dmgH),py] - gives the y-methylallylcobaloxime, the former reactions occurring by an SN2' pathway, and the latter by an SN2route. These studies demonstrate that the cobaloxime nucleophile is more sensitive to stetk effects than moxe canuent\ona\ nuc\eoph\\es such as the akoxide ion.

C\ Et-C-CZCH I

I

Me

+ [Co(dmgH),py]

~

-

H

I

+

l ~ ~ ~ C = C = C/ - C o ~ . ~ ~ H ~ * . y(


31

Stereochemical investigation of the reaction between vinyl halides and [Co(dmgH),L] (L = py, H,O or PhNH,) indicate an addition-elimination mechanism (Van Duong and Gandemer, 1970). The reactions of the cobaloximes with cis- or trans-2-phenylethenylbromide give the cis- and trans-vinylcobaloximes respectively. REACTIONS OF COORDINATIVELY-UNSATURATED COMPLEXES

A study of the reaction between trans-[IrCl(CO)(PMe,Ph),] and alkyl halides has demonstrated different behaviour to that observed with [Co(dmgH),py]-, but the details of the reaction are still unclear (Pearson and Poulos, 1979). The complications associated with the stereochemistry of the addition to the metal will be left until Section 8 and we will concentrate here on the stereochemistry observed at the carbon centre.

+

CI \ I r / ‘

P’ CH2=CHCHCICH3

a’

(75%)

CI

P l’

B CI

‘co

(25%)

/

PhMe2P

\

I\ co

CI

+

Ir

‘PMe2Ph

Scheme 10

The reaction of a-methylallyl or crotyl chloride with trans-[IrCl(CO) (PMe,Ph),] gives mixtures of products as shown in Scheme 10, the product distribution being different for the two substrates, which rules out a common

32


intermediate. This has been interpreted in terms of a mechanism involving a two-step S,2 process. The nucleophilic iridium atom attacks the saturated carbon centre, but some assistance is gained by mbonding of the double bond to the metal. This gives rise predominantly to the same o-allyl-product in which the primary carbon is bonded to iridium. It is conceivable that in methanol some x-allyl-complex formation occurs to account for the ca 20% isomerisation that is observed. However, it is difficult to rationalise the on the complete racemisation of optically active H,C==CH-CH(CI)Me basis of this mechanism. It has been proposed that racemisation is induced by the iridium complex, within the encounter complex, prior to the “oxidative addition”. 8 Stereochemical changes at the metal

The reactions of metal nucleophiles containing a coordinatively-saturated reaction centre are relatively trivial, and invariably occurs with little gross stereochemical change to the complex, as for instance in the reactions of [M(dmgH),L]- (M = Co, L = py; M = Rh, L = PPh,) or [M(CO),]( M = Mn or Re). with MeI, an almost In the reaction of [(q5-C,H,)Rh(PPh,)(q2-C,H4)] colourless solid is formed initially which is believed to be the pseudo-fourcoordinate cationic intermediate shown in Scheme 11. This complex could not be isolated in a pure state, possibly because of the extreme lability of the coordinated ethylene, but the analogous [(q 5-C,H,)Rh(AsPh,)(q2-C2H4)Me]+ has been isolated and fully characterised (Oliver and Graham, 1971). Studies on [(q5-C,H,)M(CO)(R)(I)] (M = CO or Rh, R = CF,, C,F, or C,F,) by ‘H and I9F nmr spectroscopy have demonstrated that the metal in these complexes is an asymmetric centre (McCleverty and Wilkinson, 1964).

Scheme 11

In contrast to the studies with [(q5-C,H,)Rh(PPh,)(q2-C,H4)], where predissociation of a ligand is not essential for this species to react with alkyl halides, the reactions of [Pd(PPh,),] involve the 14-electron species [Pd(PPh,),]. This behaviour is manifest in the kinetics (Stille and Lau, 1977

33

THE NUCLEOPHlLlClTY OF M E T A L COMPLEXES

and references therein; Fauvarque and Pfluger, I98 1). Similarly studies of [Pt(PPh,),(C,H,)] with MeI, PhCH,Br or ICH,CH,I shows an inverse dependence on the concentration of ethylene demonstrating that the “active species” is [Pt(PPh,),] (Birk et al., 1968). However, other studies have shown that both [Pt(PPh,),] and [Pt(PPh,),] can react (Pearson and Rajaram, 1974). Similarly, in the reactions of trans-[IrCI(CO)(PPh,),] with aryl iodides, the rate constant (/cobs) for disappearance of Ir(1) species obeys the two term rate-law shown in (26). The second term, exhibiting an inverse dependence on the concentration of PPh,, is attributed to a highly reactive, 14-electron species [IrCI(CO)(PPh,)] (Mureinik et al., 1979). In (26), K is the equilibrium constant for dissociation of PPh, from the complex, k , is the rate constant for nucleophilic attack of [IrCI(CO)(PPh,)] on ArI and k , refers to the solvent mediated pathway.

An interesting, if somewhat anomalous, conclusion comes from a kinetic study of the reactions of [RhCI(PPh,),] with H,, 0,, C,H, and MeI. Although H,, 0, and C,H, react with the species [RhCI(PPh,),], Me1 reacts not with this species, but rather [RhCI(PPh,),], and the derived dimer shown in (27). There is no apparefit reason for these various reactivities (Ohtani et al., 1977).

[RhCI(PPh,),l

+[(PPh,),Rh

c1 ’ ‘a’‘

Rh(PPh,)J

+ ZPPh,

(27)

In considering the reactions of coordinatively-unsaturated complexes, the stereochemistry of the addition at the metal can be either cis or trans. Most of the stereochemical studies in this area have been conducted with complexes of the type, trans-[IrX(CO)(PR,),]. The reaction of trans-[IrCI(CO)(PPh,),] with Me1 gave exclusively product [6], in which the substrate molecule has added trans across the plane of the complex. Furthermore, in the presence of an excess of Bu,N+Cl- the

ph3a Me

oc

1

[61

PPh,

34


exclusive product is still that shown in [6]. It was proposed that these results were a consequence of a one-step, concerted addition, consistent with theoretical considerations (Pearson and Muir, 1970). The trans-addition of alkyl halides to trans-[IrCI(CO)(PMe,Ph),l has been verified (Deeming and Shaw, 1969) and shown to yield the product whose stereochemistry is shown in [6] when the solvent is benzene. However, in methanol, a mixture is formed which comprises: [IrI,(Me)(CO)(PMe,Ph),], 55%; [6], -40%;

-

phMe2 Me

oc

c1

PMe,Ph

-

[7], 5%. This difference in the product distribution for the two solvents has been rationalised by the pathways shown in Scheme 12. Me

-

Me

+

+

+

CI,

oc/

oc‘

‘Ir’

I

L

CI

T I‘’

IL‘

oc/

I

I

CI

L

I I

I Me1

L‘

I

1

CI

I

/I

L ,

OCY1‘\

Me

Me

L

L‘

Ic1-

Ic

(A)

1-

Me

(L = monotertiary phosphine)

oc’

L‘I

I Scheme 12

In benzene, nucleophilic attack by the complex on Me1 yields the ion-pair, and subsequent attack by iodide at the metal yields the trans-product.


35

However in the more polar methanol the ion-pair can separate, and subsequent rapid reaction of iodide with trans-[IrCI(CO)(PMe,Ph),] results in halide exchange and hence the corresponding mixture of products. CH,CH= CH,

L

, \

oc

L

CI

/

CYCH=CH,

rextd

C3H5X

.

Ir\

Ir L‘

L

X

+ (L = monotertiary phosphine) L

Scheme 13

The important role of the so. Jent in defining t.: stereochemistry of the addition is further emphasised by the reactions of trans-(IrCl(C0) (PMe,Ph),] with ally1 halides (Deeming and Shaw, 1968a). Thus in benzene the product is formed as a consequence of cis-addition as shown in Scheme 13. Recrystallisation from ethanol yields the trans-product via the Ir-bonded species (which can be isolated as the BPh,- salt). Again this isomerisation appears to be the consequence of the lability of the halido-groups in the protic solvent. A mixture of products is formed in the reaction between trans-[IrBr(CO)(PMe,Ph),] and C,H,Cl because of the rapid exchange shown in (28), (Deeming and Shaw, 1968b). Consistent with these observations, when the reaction between trans-[IrBr(CO)(PMe,Ph),] and C,H,Cl is performed in the presence of an excess of LiBr, only one product is obtained, which contains no Ir-CI bonds (Pearson and Poulos, 1979).

trans-[IrBr(CO)(PMe,Ph),l

+ C1-

trans-[IrCI(CO)(PMe,Ph),]

+ Br(28)

36


The addition of acyl halides to trans-[IrCI(CO)(PMe,Ph),] also occurs with a trans-stereochemistry as shown in (29) (Deeming and Shaw, 1969; Kubota and Blake, 1971; Collman and Sears, 1968). ~,crri.s-[lrCI(CO)(PMe~Ph)~] + RCOBr

-

/-,'y

R,C40 PhMe,P

oc

(29)

PMe2Ph

Br

Further studies on these types of reactions, but with the more stericallydemanding PMe,Bu' ligand have shown that trans-addition of alkyl and acyl halides and cis-addition of allyl chloride are still the exclusive reaction modes (Shaw and Stainbank. 1972), although trans-[IrCI(CO)(PBu'R,),] (R = Et, Pr" or Bun) have a reduced tendency to react, and only do so with Mel. The reactions of [MCI(PMe,Ph),] (M = Rh or Ir) with acyl chloride yield the trans-product. However the trans-chloro-group is labile and the derived PF, - salts yield either an equilibrium mixture of the five-coordinate complex and the alkyl complex (M = Rh) or the alkyl complex exclusively (M = Ir), as shown in Scheme 14 (Bennett et al., 1981). R\ 40

C

Cl

co

CI

Scheme 14

An interesting study of the reactions of [PtMe,(diars)] {diars = 1,2-bis(dimethy1arsino)benzene) with alkyl and acyl halides has shown that here too the acyl halide adds trans, but it is not possible to decide the stereochemistry of the addition of MeI. Ally1 halides add cis, but unlike the iridium systems discussed before, the product isomer does not convert to the trans-form even under forcing conditions (Cheney and Shaw, 1971a). The reactions of [PtMe,(meso-dias)] (dias = PhMeAsCH,CH,AsMePh) with acetyl chloride, methyl iodide or allyl halides gives one isomer (Cheney and Shaw, 1971b). For acetyl chloride and methyl iodide the addition is trans, but it is impossible to distinguish between the isomers [8] and [9] on the basis of ir and nmr spectroscopy. With allyl halides the addition is cis, but again it is not possible to distinguish between the two isomers.

37


The reactions of allyl halides with [Rh{P(OMe),},]+ gives rise to mixtures of [RhX(q3-C,H,){P(OMe)3}3]+ and [Rh(q3-C3H,){P(OMe)3}4]zf, and this has been rationalised by a mechanism involving the initial cis-addition of allyl halide and subsequent displacement of either a phosphite or halide ligand, respectively (Haines, 197I ) . F, B

F2

B '0

'0

0 '

Et

I

'0

R

Et

Et

Et

[I01

In contrast to the studies with the iridium-phosphine complexes, the very reactive complex shown in [lo] reacts with alkyl halides as shown in (30), but in the presence of a large excess of LiCl the reaction of Bu"Br yields the chloro-complex under conditions where the corresponding bromo-complex does not exchange. These observations, together with the isolation of the intermediate, ~~U~~-[R~(M~)(E~,~~BF,)(NCM~)]~BF; and the reactivity order with respect to the alkyl group (Me > Et > secondary alkyl > cyclohexyl), supports an S,2 mechanism (Collman and MacLaury, 1974).

PMe,

+ C,H,, The reactions of epoxides with [IrCl(C,H,,)(PMe,),] or [Ir(PMe,),] yield the cis-hydridoalkyl-complexes shown in (3 I), where the stereochemistry of the products has been established (when R = H) from both 'H nmr spectroscopy and a crystal structure determination (Milstein, 1984; +

38


Milstein and Calabrese, 1982). However, (when R = H) there is also formed a small amount (ca 5 % ) of another isomer in which the trans-chloride and a cis-phosphine's positions are interchanged. Finally the reaction of Na,[Fe(CO),] with Pr"Br gives [Fe(CH,CH,CH,)(CO),]-, which a crystal structure determination shows has a trigonal bipyramidal structure with an apical alkyl-group (Huttner and Gartzke, 1975). The 13C nmr spectrum, however, shows only a single signal, indicating a rapid scrambling process (Collman et al., 1977). 9 The iodide catalysis effect

An aspect of the metal's coordination environment and nucleophilicity is the influence that iodide ion can have on the reactions of certain transition metal nucleophiles with MeI. This is an aspect of particular relevance to the homogeneously catalysed carboxylation of methanol to acetic acid which employs a rhodium iodide-promoted catalyst (Forster, 1979 and references therein). Kinetic studies on the iodide-catalysed reaction (32) demonstrate that in methanol an equilibrium between [Ir(cod)(phen)] and iodide (which can be studied spectrophotometrically) is established rapidly, and that both species +

[Ir(cod)(phen)]

+

+ Me1

-

[Ir( Me)( I)(cod)(phen)] +

(32)

shown in Scheme 15 react with Me1 to yield the alkylated product, the liberated iodide sustaining the catalysis. The iodo-complex reacts with Me1 about seven times faster than the parent cation, and this is presumably because of the greater electron density at the metal centre of the former (de Waal et al., 1982).

+

+

Scheme 15

THE NUCLEOPHILICITY OF

METAL COMPLEXES

39

Similarly the reactions of trans-[RhI(CO)(MPh,),1 (M = P, As or Sb) are catalysed by iodide (Forster, 1975) and the equilibrium constants for (33) K

trans-[RhI(CO)(MPh,),]

+ I - 5 [Rh(I),(CO)(MPh,)]- + MPh,

(33)

-

have been measured: M = P, K < 3 x M = As, K 0.05; M = Sb, K 2 x A lower limit for the relative rates of reactions of the anion vs the neutral species with Me1 is 1 x lo5, indicating the great increase in nucleophilicity imparted to the metal by the substitution. Semi-quantitative studies on the reaction of cis-[RhI,(CO),]- with Me1 have shown that although increasing amounts of Bu,N+I- cause a small rate-increase, this is attributable to general salt effects, but Ph,As+I-, Ph,As+CI- and Me,im+I- (Me,im = 1,3-dimethylimidazolium)cause a dramatic increase in the rate (Hickey and Maitlis, 1984). This has been interpreted in terms of the equilibria (34), where L = Me,im. A low

-

[Rh,IL(CO),I

+ I-

[Rh(I),(L)(CO),]-

=F=

[RhI,(CO),]

+L

(34)

concentration of [Rh(I),(L)(CO),]- (which may have been detected by FT ir spectroscopy) is probably the very active nucleophile.

10 The reactions of binuclear complexes

The stereochemistry of the addition of a,y-dihalides to [{(~5-C,H,)Co}, (pCO),]- has already been discussed (Section 6 ) . There are relatively few binuclear systems in which the metal centres act as a nucleophile. The reaction of [Pt,(p-dppm),] { dppm = bis(dipheny1phosphino)methane} with Me1 in benzene leads to the methylated product as shown in (35).

[Pt,(p-dppm),]

+ Me1

-

Ph2PAPPh,

I

I I

Me-Pt-Pt-PPh2-,

I

P h PVPP

h

PPh,

I

+ I-

(35)

Subsequent reaction of the methylated complex with an excess of Me1 in dichloromethane yields the dimethylated product as in (36). The binuclear system appears to be more reactive than mononuclear analogues such as [Pt(PPh,),], and this has been ascribed to anchimeric assistance by the second platinum centre (Azam et al., 1984). A similar reaction has been described for [Pd,(p-dppm),] (Balch et al., 1981).

40


Ph P,A

I

Me-Pt-Pt-PPh,

,I

P h Pv

PPh

1 I

PPh

A

,

7 PPh,

PhZP-PPh, (36)

Direct comparison of the reactivity of a mononuclear and binuclear system is possible in the reactions of [PtMe,(SMe,),] and [Pt,Me,(p-SMe,),] respectively with Me1 (Puddephatt and Scott, 1985). The dimer reacts with Me1 in acetone some twenty times faster than the monomer at - 10°C. Furthermore, in this study, the reaction of the monomer with Me1 to yieldfuc-[PtIMe,(SMe,),] in C2H3CNwas shown to proceed via the intermediate fuc-[PtMe,(SMe,),(C2H3CN)]I which was detected by low-temperature nmr spectroscopy.’ The detection of this intermediate is consistent with an S,2 mechanism. As observed in the work on [Pt,(p-dppm)3],“oxidative-addition” at one metal centre deactivates the other because of electronic factors: the metal centre is now formally in a higher oxidation-state which thus attracts charge. In the absence of restraining ligands this leads to metal-metal bond formation [see (35)]. T o circumvent these problems and to investigate the reactivity of a binuclear complex containing “isolated” metal centres the complexes shown in [ I I] and [I21 (M = Rh or Ir) have been prepared (Schenk et ul., 1985).

[ I 11

[I21

For species [ 12, X = PPh,] the reactions with Me1 and CH3COCI demonstrate the greater electron-richness compared to [12, X = CI]. The reaction with Me1 proceeds according to (37); that of the rhodium complex proceeds only to equilibrium whereas that of the iridium analogue goes to completion.

’

The prefix fur- denotes that the three methyl groups occupy mutually cis positions in an octahedral structure.


41

The stereochemistry of the addition is not exclusively that shown in (37), although the reaction exhibits a reasonably high degree of selectivity as evidenced by nmr spectroscopy. Furthermore the monoalkylated intermediate associated with raction (37) has been isolated.

FIG. 6 Reaction product between Me1 and [Ir,(C,N,H,),(cod),l

The only substantiated case of 1,2-addition of an alkyl halide across two adjacent metal atoms is shown in Fig. 6 (Coleman et a/., 1982). Finally, only one example of addition of Me1 across a hetero-binuclear complex as shown in (38) has been reported (Finke et a/., 1983). This reaction is strongly regioselective, the addition occurring in the sense shown to the extent of more than 80%.

* R h-M

M e )&gMe

O(C O ) ,

+Me1

-

+ CO (3X)

11 The reactivity of the carbon centre

In this section are included those studies in which perturbation of the carbon centre has been used as a probe for the reaction mechanism. In the succeeding section, a further aspect of the reactivity problem will be discussed, namely, the reactions of am-dihalides. Several authors have noted, although sometimes on a purely qualitative basis, that the rates of the reactions of a particular nucleophile with an alkyl or aryl halide vary inversely with the C--X bond strength (i.e. I > Br > CI), as would be expected for a nucleophilic displacement reaction (Ochiai et a/., 1969; Hart-Davis and Graham, 1970, 1971; Schrauzer and Deutsch, 1969;

42

SUNDUS HENDERSON AND RICHARD A HENDERSON

Fitton and Rick, 1971; Ramasami and Espenson, 1980; Blum and Weitzberg, 1976; Collman et al., 1977; Douek and Wilkinson, 1969). The influence of the alkyl group on the reactivity has been shown to decrease with increasing chain-length. Thus in the studies of the reaction between vitamin B,, and alkyl halides, the relative rates in methanol at 25°C are: MeCl(180) > EtCl(1.7) > P r W (1.3) > B u W ( 1 .O). The close similarity between these relative rates and the analogous reactions of iodide (although the latter were studied in acetone at 30°C) was considered indicative of the nucleophilic character of the former reactions (Schrauzer and Deutsch, 1968). The surprisingly large difference between the reactivity of the methyl and ethyl halides both in the above study and in the reactions of [(q5-C5H5)Ir(CO)(PPh,)], where Me1 reacts 400-1200 times faster than EtI, has been ascribed predominantly to the steric restrictions of the ethyl group (HartDavis and Graham, 1970). Irrespective of the leaving group, the reaction rates of alkyl compounds with various nucleophiles are perturbed by similar amounts with increasing congestion at the carbon centre. Substitution of one or two hydrogen atoms for methyl groups at the reaction centre retards the rate of the reaction with vitamin B,,, whereas increasing the carbon chain length by similar amounts has little effect (Schrauzer and Deutsch, 1969). If however the hydrogen atoms are replaced by a group which can delocalise the incipient electron density on the metal, thus increasing the stability of the transition state, the reaction rate is increased. Such groups include CN, MeO, Ph, C,,H,. H,NC=O and CO, (although, in the last case, stabilising electronic effects are overridden by the unfavourable negative charge). Consistent with the sensitivity of these reactions towards steric effects, the replacement of one hydrogen atom, at the reaction centre, by a phenyl group leads to an increased reaction rate with vitamin B,,, but substitution by two phenyl groups leads to a retardation of rate. Similar steric effects have been observed in the reactions of [Rh(dmgH),(PPh,)]with PhCH,CI ( k = 1.04 x 103dm3mol-' s - ' in aqueous methanol at 25°C) and PhCHMeCl ( k = 96 dm3 mol-' s - I ) (Ramasami and Espenson, 1980). In general the reactivity order observed with various alkyl halides and metal nucleophiles is: primary > secondary > neopentyl 9 adamantyl. The fact that adamantyl bromide is unreactive towards [Fe(CO),]'- has been used as evidence that the reactions of this dianion d o not proceed by an electrontransfer mechanism (Collman et af., 1977). In the reactions of aryl halides with various complexes, the effect of para-substituents in the benzene ring on the reaction rate has been used as a probe of the reaction pathway. The reaction of [Pd(PPh,),J with p-RC,H,I as shown in Scheme 16 is influenced by the substituent; electron-withdrawing groups enhance the rate in the order: NO, > CN > PhCO > H. This


43

Scheme 16

order indicates that the reaction is, in essence, an aromatic nucleophilic substitution, where breaking of the bond to the leaving group is rate-limiting (Fitton and Rick, 1971). However, the same reactivity pattern could result from an electron-transfer mechanism. A similar study (Fauvarque and Pfluger, 1981) demonstrated a good correlation of log,,k, with Hammett o ( p = 2.0). This p-value is consistent with an increased electron density transferred to the aromatic ring in the transition state. The reaction proceeds most readily with aryl iodides, and this has been attributed to the halogen being used as a ligand in the intermediate/transition-state as shown in [13].

Although in these reactions between [Pd(PPh,),] and aryl halides, no products typical of radical reactions (e.g. ArH or ArAr) were observed, similar reactivity constants (p 2.0) have been observed in the reactions of [NiBr(PPh,),] with aryl halides, and this has been interpreted in terms of an electron-transfer reaction (Tsau and Kochi, 1979). In a study of the reactions of [Pt(R),(bipy)] in acetone solution, the position of the electronic transition of lowest energy in the uv/visible spectrum, which is a metal-to-ligand charge-transfer band (MLCT), was found to correlate well with the value of log,,k, (where k, is the secondorder rate constant for the reaction with MeI). The lowest energy of the MLCT band occurs when the highest electron density is at the metal. The correlation thus strongly suggests that the rate of the reaction is primarily dependent on the energy of the filled d-orbitals on platinum. The value of k, varies from 2.8 x lo3 to 2.3dm3 m o l - l s - l , in the order: R = Me > p-MeOC,H, > p-MeC,H, > rn-MeOC,H, > C,H, > p-FC,H, > p-ClC,H, > rn-CIC,H, (Jawad and Puddephatt, 1977). Interestingly, the

-

44


position of the MLCT band for R = p-MeC,H, is in about the same position as for o-MeC,H,, but this complex does not react with Me1 presumably because of the unfavourable steric interactions. The Hammett correlation with values of k, is not very good, presumably because specific solvation effects, which are reflected in both the MLCT band and the value of k,, are not represented in the Hammett o-values. The reaction constant for [Pt(Ar) ,(bipy)] with Me1 per aryl group (p - 1.3) is very similar to that per aryl group for rrans-[IrCl(CO)(PAr3),1 (p = - 1.1) (Ugo et al., 1972). If such a comparison is valid, it indicates that, despite the very different distances of the aryl groups from the reaction centre in the two types of complex, very similar electronic effects are transmitted. The negative values of the reaction constants demonstrate that positive charge is formed on the metal atom in the transition state. In studies on trans-[IrCl(CO)(PAT&] with alkyl halides (Ugo et al., 1972) the value of the reaction constant for Me1 (-6.4) and PhCH,CI (-2.6) were obtained. The less negative value for the reaction constant with benzyl chloride is attributable to delocalisation of the negative charge (in the transition state) over the benzene ring.

-

12 The reactions of a,o-dihalogenoalkanes

The reactions of a,w-dihalogenoalkanes with transition-metal nucleophiles proceed by a variety of different routes. In the simplest possible pathway only one end of the alkyl-group is substituted (39); M = M o or W, X = Br [(‘I’-C,H,)M(CO)J

+ X(CH,),X

-[(‘15-C,H,)M{(CHz),X}(CO),I

+ X(39)

or I (King and Bisnette, 1967). However, in the presence of an excess of the metal complex, subsequent replacement of the remaining halide results in a dimetallo-species as shown in (40); M = Fe or R u (King, 1963). Xrl ‘-C,H ,)M(CO),I

+ Br(CH,),Br

-

[{(‘I 5-C,H,)M(CO)z}z(CH,),]2 +

+ 2Br (40)

Alternatively, after the initial alkylation step, subsequent attack by the excess of metal complex on the metal alkyl species results in migration of the

Scheme 17

Br


45

alkyl group to an adjacent carbonyl ligand (Scheme 17). Subsequent intramolecular attack of the carbonyl oxygen on the alkyl halide yields a carbene ligand (Casey and Anderson, 1971). The structure of the carbene complex has been established by X-ray crystallography. A second metal centre is not essential to yield the carbene complexes, however. Thus the reaction of [(q5-C5H5)M(CO),]- (M = Mo or W) with 1,3-diiodopropane in T H F or 1,2-dimethoxyethane at reflux yields the cyclic carbene complex, presumably

Scheme 18

by the mechanism shown in Scheme 18 (Adams et al., 1984 and references therein). In the presence of a large excess of [(q5-C,H,)Mo(CO),]-, compound [14] is produced, presumably by a pathway analogous to that shown in Scheme 17.

Q

co 1 ,.co -MO’ 1 ’co

~ 4 1

This proposed mechanism is further reinforced by the formation of the heterobimetallic species of the structure shown in [ I 51 formed from the reaction between [(q’-C,H,)W(CO),]and [(q5-C,H,)Mo{(CH,),Br}(CO),]. However, here the carbene is unexpectedly coordinated to the tungsten atom. A similar transfer of the ligand, proceeding after the alkylation but prior to the ring-closure, occurs in the reaction of [Mn(Me)(CO),] with [Re(CO),]- (Casey et al., 1975).

Poo

oc.... oc‘ A.

I0 C w \% ’

~ 5 1

46


The reactions of [(q5-C,H,)M{(CH2)X}(CO),] (M = Mo or W, X = Br or I) with transition metal nucleophiles is not a general method for making heterobimetallic carbene complexes, since the products can be homobimetallic species involving displacement of the originally alkylated metal centre as shown in Fig. 7. -

X+

-X -

Fe OC/

\ co

OC"i0\0 oc

I\

oc

co

FIG. 7 Reaction between [(q5-C5H,)Mo{(CH2),X}(CO),] and an excess of [(T15-C5H5)Fe(Co),l-

The reactions of 1,1, I -tris(halogenomethyl)ethanes, MeC(CH,X), (X = Br or I), with the nucleophiles [(q5-C5H5)Fe(CO),)]-, [(q5-C,H,)M(CO),]- (M = M o or W) [(q'-C,H,)Ni(CO)] or [Re(CO),]all give the 1 -methylcyclopropylmethyl-derivatives as shown in Fig. 8. No

Me I

FIG. 8

Reaction between [(q5-C5H5)W(CO),]- and CH,C(CH,Br),


47

reaction occurs with [Co(CO),] -, presumably because of its poor nucleophilicity. Several mechanisms are consistent with the formation of these products including an electron-transfer process and two alternative nucleophilic pathways shown in Scheme 19 (Poli et al., 1985). It is currently impossible to distinguish between these possibilities.

Scheme 19

.

The reaction of [Re(CO),] - with MeC(CH,I), also produces the 1 -methylcyclopropylmethyl-derivative, but [Re,(CO),I] - is also formed in appreciable amounts. In contrast, [Mn(CO),]- reacts with MeC(CH,I), to yield the “carbonyl inserted” acyl-product as shown in (41). Mechanisms have been postulated to rationalise these products but these are highly speculative. 4[Mn(CO),]-

+ MeC(CH,I),

-

[Mn,(CO),,]-

+

(CO),Mn-C

+ 3113 Activation parameters

%I1

(41)

Within this section will be discussed the activation parameters ( A H * , AS* and AV*) that have been measured for the reactions between metal nucleophiles and carbon compounds. The values of the enthalpy and entropy of


48

activation for the reactions of many metal nucleophiles with Me1 and MeOTs have been summarised (Pearson and Figdore, 1980). The parameters for the reactions of alkyl halides with metal complexes show a great deal of uniformity: A H * 10 _+ 5 kcalmol-' and A S * -30 _+ 20 cal K-' mol- '. However, such values are insufficiently diagnostic to allow discrimination between the complex acting as a nucleophile or as a reductant towards alkyl halides (see Section 2) (Pearson and Figdore, 1980). More recent data also contain activation parameters in the range indicated above. Thus the reaction of [Fe(CO),(PMe,),] with Me1 has A H * = 13 kcal mol-', and AS* = -33 cal K - ' mol-' (Bellachioma et al., 198 I ), and the reactions of [Ir(cod)(phen)J+and [Ir(cod)(phen)I] with Me1 are associated with the parameters, A H * = 12.4 kcalmol-', A S * = -34.9 cal K - ' mol-', and A H * = 10.7 kcalmol-', A S * = -38.0 cal K-lrnol-' respectively (de Waal, 1982). The solvent has a relatively large influence on the activation parameters as demonstrated in the reaction of cis-[RhI,(CO),]- with Me1 (Hickey and Maitlis, 1984). For the range of solvents, methanol, chloroform, T H F and methyl acetate the values of A H * progressively decrease from 16.5 to 11.2 kcal mol-', and AS' decreases from -22.0 to -42.1 cal K - ' mol-'. The values of the activation parameters, together with the influence of the solvent on these parameters, has been interpreted in terms of a highly linear, polar transition state (Fig. 2) (Halpern, 1970 and references therein), or a transition state containing unusually stringent stereochemical restrictions. These conclusions are further substantiated by the comparable reactivities of substituted benzyl bromides towards both trans-[IrCl(CO)(PPh,),] and tertiary amines. The large negative values of A S * in the reactions of the iridium(1) complexes have been rationalised in terms of increased solvation of the transition state, attributable to its increased dipole. Such a dipole results not only from the interaction of the alkyl halides with the metal centre, but also from deformation of the iridium complex from planar to pseudo-octahedral geometry as would occur in a three-centre cis-addition (Harrod and Smith, 1970). In some cases the values of A S * and the influence of solvent on A H * and A S * have been explained in terms of a polar, asymmetric, three-centre transition state, in which the interaction of the metal is predominantly with the carbon centre (Ugo et al., 1972). The characteristic values of A H * and A S * have been taken to indicate that a nucleophilic displacement reaction is rate-limiting in a multistep reaction. For instance reaction (42) occurs in two stages, but the intermediate which is detected is too reactive to be isolated and characterised. The

-

[RhCI(CO),(PPh,)]

-

+ Me1

-

[Rh(CI)(I)(CO)(COMe)(PPh,)] (42)


49

first-order dependence of the kinetics of its formation on the concentration of Me1 and the activation parameters, AH* = 1 1.5 kcal mol- ', AS* = - 33 cal K - ' mol- ', indicate nucleophilic attack of [RhCI(CO),(PPh,)] on Me1 to yield the intermediate [Rh(CI)(I)(Me)(CO),(PPh,)]. Subsequent intramolecular migration of the alkyl group on to a carbonyl carbon atom yields the acyl product (Uguagliati et al., 1970). One apparent nucleophilic reaction which has somewhat anomalous activation parameters is that of [Pd(PPh,),] with aryl iodides, where AH* = 18.4 kcalmol-' and AS* = + 3.1 calK-'mol-'. These values, however, may be a consequence of a slightly different transition state involving the assistance of the iodine atom acting as a ligand as shown in [ 131 (Fauvarque and Pfluger, 1981). There has been only one report to date of the determination of the volume of activation (AV") in a reaction involving a metal nucleophile and alkyl halide (Steiger and Kelm, 1973). The measured volume of activation is made up of two terms, one involving the change in volume during the formation of the transition state (A V:) and the other involving the accompanying change in solvation (AV:). The latter value can be estimated using the pressure derivative of the Kirkwood formula. Study of the reaction between trans[IrCI(CO)(PPh,),] and Me1 in a range of solvents gave AV; = 17 cm3 mol-', a value typical of a bimolecular reaction (for the Menschutkin reaction between Me1 and pyridine. A V: = - 22 cm3 mol- '). The value is rather small for the simultaneous formation of two bonds (three-centre transition state), however, and is more consistent with a linear transition state.

14 Thermodynamics of reactions involving metal nucleophiles

There has been relatively little work on the thermodynamics of the reactions between transition-metal nucleophiles and alkyl halides. However, the enthalpies of the reactions between trans-[IrC1(CO)(PMe3),] and a variety of alkyl and acyl halides have been measured using titration calorimetry in dichloromethane (Yoneda and Blake, 198I). The - AH"-values calculated are: MeI, 28.1 kcal mol-'; EtI, 25.6; Pr"1, 24.4; PriI, 21.1; PhCH,I, 22.7; MeCOI, 30.0; PhCOI, 29.0 (all values refer to the alkyl iodide in its standard state). Using these data, information about the iridium-ligand bond strengths were obtained, with the assumption that the difference between the heats of sublimation of reactants and products is approximately zero. In this way the iridium-arbon bond strengths were shown to fall in the order: CH, CH,CO Pr" Et Pr' > PhCH,, the same order as observed in the corresponding alkanes and alkyl iodides.

-

- - -

50


A further calorimetric study on the reactions of [Pt(PPh,),(C,H,)] with Me1 gives AH” = - 18.9 kcal mol-’. From the corresponding studies with iodine and 1,2-diiodoethane (to give [Pt(PPh,),I,] in both cases), and assuming that the enthalpies of sublimation of [Pt(PPh,),(C,H,)], [Pt(PPh,),I,] and [Pt(PPh,),(Me)I] are all approximately the same, then it can be calculated that D(Pt-Me) - D(Pt-I) = 1.4 1.2 kcal mol-’. The platinum4arbon and platinum-iodine bond strengths are thus about the same (Mortimer et al., 1979). 15 Activation of carbon-hydrogen

bonds

A review on the nucleophilicity of metai complexes would not be complete without some mention of the reaction of alkanes and arenes with transitionmetal complexes. As recently pointed out (Halpern, 1985), a mechanism for such reactions involving the nucleophilic displacement of H - is energetically too demanding. A much more reasonable pathway involves electrophilic displacement, one-centre concerted addition and two-centre addition. Clearly factors such as the degree of coordinative unsaturation and the thermodynamics of the reaction (attempting to trade-off the strength of the C-H bond for the M-C and M-H bonds) are of prime importance in activating carbon-hydrogen bonds. However, particularly in a one-metal centre mechanism, the nucleophilicity of the metal centre must play a significant part, although an electron-rich metal centre is not essential because of the multiple roles played by the site in these reactions (Zeimer et a/., 1984). Thus the complexes discussed in this section will not be the same as those encountered so far in this review. It is only in the last decade that the activation of simple hydrocarbons by a simple transition metal complex has been realised (Crabtree et al., 1979; Baudry et al., 1980; Green, 1978; Janowicz and Bergman, 1982). Prior to this any mechanistic information about the possible pathways had to be gleaned by investigating the microscopic reverse of the “activation” process, the more commonly encountered elimination of alkane (43). Such studies can be misleading, and led to the

speculation that carbon-hydrogen activation would not be possible, on thermodynamic grounds (Sen and Halpern, 1978). With the discovery of more systems which “activate” hydrocarbons, however, mechanistic studies are beginning to appear in this area. This section is in no way supposed to represent a detailed survey of carbon-hydrogen bond activation, but concentrates only on the mechanistic studies. The literature associated with C-H and C-C bond cleavage by gas-phase transition-metal ions will not

ISOTOPE EFFECTS ON NMR SPECTRA

151

The equilibrium constant for [6], K = 1.0197 (-135°C) to 1.0114 (-62°C) was calculated from the equation K = (A +26)/(A -26) using A = 277 ppm. From the van't Hoff equation AW = -6.0 cal mol-' and AS" = -0.056 cal mol- K - . Any intrinsic isotope shifts were estimated from the precursor alcohol to be smaller than 0.06 ppm. The observed downfield shift shows that the equilibrium favours the positive charge on 13C. The vibrational frequencies associated with the charged carbon must therefore be higher than those of the methine carbon. The more confined bonding situation for the charged carbon despite the smaller number of attached atoms could indicate stiffer C-C bonding caused by C-H-hyperconjugation with the methyl groups.

'

*CH,

CH3

CH3

CH,

A secondary 13C equilibrium isotope effect was measured in the 1,2dimethylcyclopentyl cation [ I471 which was 3C-labelled in one methyl group (Saunders et al., 1977a). In the 13Cnmr spectrum of cation [I471 the carbon next to the I3C appeared as a doublet (Jcc = 35 Hz) offset downfield between 0.25 ppm ( - 125°C) and 0.10 ppm (-65°C) from the singlet of the other averaged C+/C-H carbon. This shows that the labelled methyl group is preferred next to the charged carbon. Although the C-H frequencies are lower for this methyl group as evident from hydrogen isotope effects in [116] and [ 1441, the 3C isotope effect on (104) indicates a small net increase of all vibrational frequencies of the carbon of the CH,-group at the C+-position as compared to the other CH,-group. Both '3C-equilibrium isotope effects in [6] and [147] have the same direction and may be regarded as another piece of evidence for C-H hyperconjugation. In valence bond terminology the bond between the charged carbon and the attached methyl group is stiffer because of partial double bond character. The heavier isotope (' 'C) prefers this more confined bonding, whereas the lighter isotope (12C) prefers to be at the remote position where the bonding is less stiff. Another secondary 3C-equilibrium isotope effect has been observed in [5] using natural abundance I3C nmr spectroscopy (Olah et al., 1985a).

52


*

thermodynamic preference for arene over alkane activation is also reflected in the kinetics where the difference in free energies of activation for benzene over cyclopentane is AAG* = 0.8 kcal mol-' (Jones and Feher, 1984).

Rh

-

3$r

L

FIG. 10 Competitive equilibria used to establish the discrimination between intraand intermolecular arene activation

It has often been suggested that entropy effects dominate the discrimination between intra- and intermolecular C-H activations (Dicosimo et al., 1982 and references therein). Estimates of - TAS* = 10 kcal mol-' have been made. However several diverse observations on the affinity of systems to intra- and intermolecular activation clearly require this point to be elaborated. This has been made possible by direct comparison on a single metal site illustrated in Fig. 10 (Jones and Feher, 1985). The intermediacy of an q2-arene species in the intramolecular reaction can be demonstrated by the independent preparation of [I61 which rearranges to [I71 as shown in (44). This reaction occurs at a temperature below that required for arene dissociation and so it presumably occurs via the intermediate shown. 1

I

~ 7 1

"61

The results demonstrate that the activation parameter differences for inter- )'s intramolecular reactions are only small (AAH* = 1.7 kcalmol-', AAS" = 4.5 cal K - ' mol- and this corresponds to a kinetic preference for intermolecular activation of I .86 : 1. The intramolecular product is the thermodynamically more favoured, however. An analogous comparison of

',


53

alkane activation using [(q5-C5Me5)Rh(PMe,Pr")]and propane demonstrated that, as with arenes, intermolecular activation is favoured kinetically (AAH* = 5.1 kcal mol-I) but not thermodynamically (AAH" = 4.5 kcal mol-I). Direct comparison of the reactivity and selectivity of various alkanes towards [(q5-C5Me5)M(PMe,)](M = Rh or Ir) has shown that the rhodium complex is more discriminating than the iridium analogue. The former reacts almost exclusively with primary carbon centres, whereas the latter discriminated only weakly between primary and secondary carbon atoms (Perland and Bergman, 1984). Employing this weak discrimination and measuring the free energy change associated with equilibrium (45), it is estimated that the iridium-primary carbon bond is 5.5 kcal mol- stronger than an iridiumsecondary carbon bond (Wax et al., 1984).

(45)

Clearly a great deal of work is required before we can define the way in which the metal cleaves the carbon-hydrogen bond. Recently (Crabtree et af., 1985) structural information from 18 crystal structures has been used to construct a reaction trajectory for the reaction between a metal centre and the C-H unit, in which the C-H bond approaches the metal and gradually swings and elongates the carbon-hydrogen bond to form the alkyl hydridoproduct. The importance of steric effects in C-H activation has been stressed on theoretical grounds (Saillard and Hoffman, 1984). 16 Applications

The reactions between metal nucleophiles and organic compounds finds several applications in the areas of organic synthesis and catalysis. One obvious area of application is the control of stereochemistry at the carbon centre when the metal nucleophile inverts the configuration of the carbon in an S,2 process. In this way the C-20 centre of steroids has been controlled by a double inversion sequence using [Pd(PPh,),] (Trost and Verhoeven, 1976). Clearly, the choice of nucleophile is important when stereochemically indiscriminate electron-transfer processes are so often energetically very comparable.

54


A reagent which has been referred to as a “transition-metal analogue of the Grignard reagent” is Na,[Fe(CO),] (Collman, 1975 and references therein). The versatility of this reagent is attributable primarily to its high stereospecificity and its specificity to particular groups, leaving many (such as carbonyls, esters, acids, etc.) completely unperturbed, thus circumventing the problems associated with the use of protecting groups. Alkyl halides can be converted to aldehydes (Cooke, 1970; Watanabe et al., 1971) as illustrated in (46). Such a reaction uses to advantage the tendency of the derived

(46)

alkyl complex, [Fe(R)(CO),] -, to rearrange in the presence of carbon monoxide to yield the corresponding acyl-complex [Fe(COR)(CO),] - . A similar strategy, this time quenching the reaction with alkyl halide, yields unsymmetrical ketones as illustrated in (47) (Collman et al., 1972). This reaction is not truly general, however, since the “quenching” alkyl halide must contain a primary alkyl group. -CO,E~

vco

[FdCOIJ’ . N M P

Br

0

Ell

Alkyl halides can be converted into carboxylic acids by exposing the reaction mixture to dioxygen (48), or the corresponding ester if the reaction [Fe(CO) 1’

n-C ,H, ,CI 4 n-C ,H, ,CO,H

n-C,H,CI

[Fe(CO),lz -

I,. MeOH

n-C,H,CO,Me

(49)

is quenched with an alcohol (49) (Collman et al., 1973). In an analogous fashion quenching the reaction with amines gives amides. The limitations of this reagent are associated with its basicity, which in its reactions with alkyl halides can give rise to elimination, thus prohibiting the use of tertiary alkyl halides. The other limitation of this reagent is associated with the migratory capabilities of the alkyl-group which is greatly impeded by adjacent electronegative groups on the carbon residue; thus the reactions are restricted to those of simple primary and secondary substrates. The reactions of acyl halides with [FeH(CO),]- have been used to prepare aldehydes, by the (assumed) pathway shown in (50) (Cole and Pettit, 1977;


55

this paper contains several references to other applications of this versatile reagent). R\

C , =O CI

+ [FeH(CO),]-

-

RCO H,

I ,co

Fe

-

,R C ,=O H

(50)

The mixture of [Fe(CO),] and NaOH in 95% methanol is a good, selective hydrogenator for a,p-unsaturated carbonyl compounds (Noyori et al., 1972). It is not entirely clear what the exact nature of the “active” species is, but the wine colour of the solution may suggest the formation of [HFe,(CO),]- and [HFe,(CO), ,I-. The nucleophilicity of transition metal complexes towards organic molecules has found application in industrial catalysts. The isomerisation of epoxides to ketones takes place at the complex [RhCl(PMe,),], as shown in Scheme 20 (Milstein, 1984 and references therein). In this catalytic cycle, the cis-[RhH(CH,COR)Cl(PMe,),l intermediate can be isolated, and is clearly analogous to the indium system described in Section 8.

Scheme 20

56


A variety of reactions, such as the co-catalysed hydrogenation of aldehydes and the rhodium-catalysed decarbonylation, hydroacylation, hydrogenations and hydroformylation of aldehydes may involve the “oxidative addition” of aldehydes to the metal centre. Furthermore, although the addition of aldehydes to tr~ns-[RhCl(CO)(PPh,)~] (Landvatter and Rauchfuss, 1982), [Ir(PMe,),]+ (Thorne, 1980) and [RhCI(PMe,),] (Milstein, 1982) has been shown to yield cis-hydridoacyl-complexes, it is not clear that any of these reactions involve the nucleophilicity of the metal. Certainly other means of bonding of the aldehyde (via the carbonyl oxygen) can be envisaged. Probably the most important industrial application of a transition-metal nucleophile is the Monsanto process for carbonylating methanol, using soluble rhodium-carbonyl complexes in the presence of iodide. The catalyst is in fact [RhI,(CO),]- (Forster 1979, and references therein), and the catalytic cycle is shown in Scheme 21. The substrate for the rhodium catalyst is methyl iodide, which “oxidatively adds” to yield [Rh(Me)(I),(CO),]-. 0 CH&’

Y)R

HI

+

a!: /

oc

I

-

h co

Scheme 21

I2

3

I

0


57

Subsequent intramolecular migration of the methyl group on to the carbonyl group, and presumed reductive elimination of MeCOI, regenerates the catalyst. Solvolysis of MeCOI by either water or methanol, yields acetic acid or methyl acetate, respectively (Masters, 1981 and references therein). The role of iodide in this catalyst has already been discussed in Section 9. The analogous iridium system is more complex, with two possible catalytic cycles, depending upon the reaction conditions, one involving [IrIz(CO)z]- and the other involving [IrI(CO),] as shown in Scheme 22.

CH,CO,H

co

‘J

\ [Ir(MeCO~I,(COl,l, or

Me1

oc 2

x

0

Scheme 22

,

58


The prime difference between the rhodium and iridium systems is that whereas in the former the rate-limiting step is the nucleophilic attack of [RhI,(CO),]- on methyl iodide, in the iridium system the addition of methyl iodide is rapid, but the subsequent intramolecular migration is slow. 17 Adfinern

In this review we have indicated the sort of metal complexes that react as nucleophiles and the type of organic molecules with which they react. What then of the future? Just a casual glance through many of the sections of this review reveals that many aspects of this area of chemistry remain incomplete. A constant dichotomy in the reactions of metal complexes with alkyl halides is whether the complex is truly reacting as a nucleophile or as a reductant, and a facile discrimination between the two roles has not been found as yet. An area of great interest currently is the “activation” of C-H and C-C bonds by simple transition-metal complexes. The delineation of the factors involved in these stoichiometric reactions, in particular to what extent the nucleophilic character of the complex plays a role, is of fundamental importance in designing a catalytic system. In Section 4 the problems associated with establishing a nucleophilicity scale for metal complexes was discussed; we must await, with some impatience, the establishment of such a scale.

References

Adarns, H.. Bailey, N. A. and Winter. M. J. (1984). J . Chem. SOC.Dalton Trans. 273 Alper. H. (1975). Tef. Left. 2257 Anderson, S . N., Ballard. D. H., Chrzastowski, J. Z., Dodd, D. and Johnson, M. D. ( I 972). J . Chem. SOC.Chem. Cornmun. 685 Ashby. E. C. and DePriest, R. (1982). J . Am. Chem. SOC.104, 6144 Azam, K . A., Brown, M. P., Hill, R. H., Puddephatt, R. J. and Yavari, A. (1984). Orgnnome~allics3, 697 Balch, A. L. Hunt, C. T. Lee, C. L., Olrnstead, M . M. and Farr, J. P. (1981). J. Am. Climi. Soc. 103, 3764 Baudry, D.. Ephritikhine, M . and Felkin, H. (1980). J. Chem. SOC.Chem. Commun. 1243 Becker, Y . and Stille. J. K . (1978). J . Am. Chem. SOC.100, 838 Bellachiorna, G.. Cardaci, G and Reichenbach, G. (1981). J . Organomet. Chem. 221, 29 1 Bennett, M . A., Jeffery, J. C . and Robertson, G. B. (1981) Inorg. Chem. 20, 323 Birk. J. P.. Halpern. J. and Pickard, A. L. (1968). J. Am. Chem. SOC.90, 4491 Block. P. L. and Whitesides, G. M. (1974). J . Am. Chem. SOC.96, 2826


59

Block, P. L., Boschetto, D. J., Rasmussen, J. R., Deavers, J. P. and Whitesides, G. M. (1974). J . Am. Chem. SOC.96, 2814 Blum, J. and Weitzberg, H. (1976). J . Organomet. Chem. 122, 261 Booth, B. L., Haszeldine, R. N. and Perkins, I. (1975). J. Chem. SOC.Dalton Trans. 1843 Bradley, J. S., Connor, D. E., Dolphin, D., Labinger, J. A. and Osborn, J. A. (1972). J. Am. Chem. Soc. 94, 4043 Calderazzo, F. and Noack, K. (1966). Coord. Chem. Rev. 1, 118 Casey, C. P. and Anderson, R. L. (1971). J . Am. Chem. SOC.93, 3554 Casey, C. P., Cyr, C. R., Anderson, R. L. and Marten, D. F. (1975). J . Am. Chem. Soc. 97, 3053 Chatt, J. and Davidson, J. M. (1965). J . Chem. SOC.843 Cheney, A. J. and Shaw, B. L. (1971a). J . Chem. SOC.( A ) 3545 Cheney, A. J. and Shaw, B. L. (I971b). J. Chem. SOC.( A ) 3549 Chock, P. B. and Halpern, J. (1966). J . Am. Chem. SOC.88, 35 1 1 Chock, P. B. and Halpern, J. (1969). J . Am. Chem. SOC.91, 582 Cole, T. E. and Pettit, R. (1977). Tet. Lett. 781 Coleman, A. W., Eadie, D. T. and Stobart, S. R. (1982). J . Am. Chem. SOC.104, 922 Collman, J. P. (1975). Acc. Chem. Res. 8, 342 Collman, J. P. and MacLaury, M. R. (1974). J. Am. Chem. SOC.96, 3019 Collman, J. P. and Roper, W. P. (1968). Adv. Organomet. Chem. 7, 53 Collman, J. P. and Sears Jr, C. T. (1968). Inorg. Chem. 7, 27 Collman, J. P., Kubota, M., Sun, J. Y., Kang, J. and Vastine, F. (1968). J . Am. Chem. SOC.90, 5430 Collman, J. P., Winter, S. R. and Clark, D. R. (1972). J . Am. Chem. SOC.94, 1788 Collman, J. P., Winter, S. R. and Komoto, R. G. (1973). J. Am. Chem. SOC.95, 249 Collman, J. P., Finke, R. G., Cawse, J. N. and Brauman, J. I. (1977). J. Am. Chem. SOC.99, 25 15 Cooke, M. P. (1970). J . Am. Chem. Soc. 92, 6080 Cooksey, C. J., Dodd, D., Gatford, C., Johnson, M. D., Lewis, G. J. and Titchmarsh, D. M. (1972). J . Chem. SOC.Perkin Trans. 2 655 Cotton, F. A. and Wilkinson, G. ( I 980). “Advanced Inorganic Chemistry”, 4th edn, Ch. 3. Wiley Interscience, New York Crabtree, R. H. and Hlatky, G. G . (1980). Inorg. Chem. 19, 571 Crabtree, R. H., Mihelcic, J. M. and Quirk, J. M. (1979). J. Am. Chem. Soc. 101, 7738 Crabtree, R. H., Holt, E. M., Lavin, M. and Morehouse, S. M. (1985). Inorg. Chem. 24, 1986 Darensbourg, M. Y. and Hanckel, J. M. (1981). J . Organomet. Chem. 217, C9 Darensbourg, M. Y. and Hanckel, J. M. (1982). Organometallics 1, 82 Darensbourg, M. Y., Darensbourg, D. J., Burns, D. and Drew, D. A. (I 976). J. Am. Chem. SOC.98, 3127 Darensbourg, M. Y., Jimenez, P. and Sackett, J. R. (1980). J . Organomet. Chem. 202, C68 Darensbourg, M. Y., Jimenez, P., Sackett. J. R., Hanckel, J. M. and Kump, R. L. (1982). J . Am. Chem. SOC.104, 1521 Deeming, A. J. (1972). M . T.P. International Review qf Science, Ser. 1. (ed. M. L. Tobe) 9, 117 Deeming, A. J. (1974). M . T . P . International Review of Science, Ser. 2. (ed. M. L. Tobe) 9, 271

60

SUNDUS HENDERSON AND RICHARD A HENDERSON

Deeming, A. J. and Shaw, B. L. (1968a). J . Chem. SOC.( D ) 751 Deeming, A. J. and Shaw, B. L. (1968b). J . Chem. SOC.( A ) 1562 Deeming, A. J. and Shaw, B. L. (1969). J . Chem. SOC.( A ) 1 128 Deeming, A. J. and Shaw, B. L. (1971). J . Chem. SOC.( A ) 1802 Dessy, R. E., Pohl, R. L. and King. R. B. (1966): J . Am. Chem. SOC.88, 5121 de Waal. D. J. A,, Gerber, T. I . A. and Wynand, J. L. (1982). Inorg. Chem. 21, 1529 DiCosimo. R.. Moore, S. S.. Sowinski, A. F. and Whitesides, G . M. (1982). J . Am. Chem. Soc. 104. 124 Douek, I . C. and Wilkinson, G . (1969). J. Chem. SOC.( A ) 2604 Edgell, W. F. and Chanjamsri, S. (1980). J . Am. Chem. SOC.102, 147 Edgell, W. F., Hedge, S. and Barbetta, A. (1978). J . Am. Chem. Soc. 100, 1406 Fauvarque, J. F. and Pfluger, F. (1981). J . Organomet. Chem. 208, 419 Finke. R. G., Gaughan, GT, Pierpont, C. and Noordik, J. H. (1983). Organometallics 2. 1481 Fitton, P. and Rick, E. A. (1971). J . Organomet. Chem. 28, 287 Forster, D. (1975). J . Am. Chem. SOC.97, 951 Forster, D. (1979). Adv. Organomet. Chem. 17, 255 Franks, S. and Hartley, F. R. (1981). Inorg. Chim. Acta 49, 227 Franks, S., Hartley, F. R. and Chipperfield, J. R. (1981). Inorg. Chem. 20, 3238 Gaylor, J. R. and Senoff, C . V. (1972). Proc. X I V . Int. Coord. Conf Chem. Green, M. L. H. (1978). Pure Appl. Chem. 50, 27 Haines, L. M. (1971). Inorg. Chem. 10, 1693 Halpern, J. (1970). Acc. Chem. Res. 3, 386 Halpern, J. (1985). Inorg. Chim. Acta 100, 41 Hanratty, M. A,, Beauchamp, J. L., Illies, A. J. and Bowers, M. T. (1985). J . Am. Chem. SOC.107, 1788 Harrod, J. F. and Smith, C. A. (1970). J . Am. Chem. SOC.92, 2699 Hart-Davis, A. J. and Graham, W. A. (1970). Inorg. Chem. 9, 2658 Hart-Davis, A. J. and Graham, W. A. (1971). Inorg. Chem. 10, 1653 Hayashi, T., Hagihara, T., Konishi, M. and Kumada, M. (1983). J . Am. Chem. SOC. 105, 7767 Hickey, C. M. and Maitlis, P. M. (1984). J . Chem. SOC.Chem. Commun. 1609 Hill, R . H. and Puddephatt, R . J. (1985). J. Am. Chem. SOC.107, 1218 Houriet, R., Halle, L. F. and Beauchamp, J. L. (1983). Organometallics 2, 1818 Huttner, G. and Gartzke, W. (1975). Chem. Ber. 108, 1373 Jawad, J. K. and Puddephatt, R. J. (1976). J . Organomet. Chem. 117, 297 Jawad. J. K . and Puddephatt, R. J. (1977). J . Chem. SOC.Dalton Trans. 1466 Janowicz. A. H. and Bergman, R. G. (1982). J . Am. Chem. SOC.104, 352 Jensen, F. R. and Davis, D. D. (1971). J . Am. Chem. SOC.93, 4048 Jensen, F. R. and Knickel. B. (1971). J . Am. Chem. Soc. 93, 6339 Jensen, F. R., Madan. V. and Buchanan, D. H. (1970). J. Am. Chem. Soc. 92, 1414 Jensen, F. R . , Madan, V. and Buchanan, D. H. (1971). J . Am. Chem. SOC.93, 5283 Johnson, M. D. (1970). J . Chem. SOC.(D)1027 Johnson, R. W. and Pearson, R. G . (1970). J . Chem. Soc. (D)986 Jones, W. D. and Feher, F. J. (1984). J . Am. Chem. SOC.106, 1650 Jones, W. D. and Feher. F. J. (1985). J . Am. Chem. SOC.107, 620 Jordan, R . F. and Norton, J. R. (1982). J . Am. Chem. SOC.104, 1255 King, R. B. (1963). Inorg. Chem. 2, 531 King, R. B. (1975). J . Organomet. Chem. 100, 1 1 1 King, R. B. and Bisnette, M. B. (1967). J . Organornet. Chem. 7, 31 1


61

Kitching, W., Olszowy, H., Waugh, J. and Dodrell, D. (1 978). J . Org. Chem. 43, 898 Kochi, J. K. (1 978). “Organometallic Mechanisms and Catalysis”. Academic Press, New York, San Francisco, London Koermer, G . S., Hall, M. L. and Traylor, T. G . (1972). J . Am. Chem. Soc. 94, 7205 Kubota, M. and Blake, D. M. (1971). J . Am. Chem. Soc. 93, 1368 Kubota, M., Kiefer, G. W., Ishikawa, R. M. and Benacia, K. E. (1973). Inorg. Chim. Acta 7, 195 Kuivila. H. G . and Alnajjar, M. S. (1982). J . Am. Chem. Soc. 104, 6146 Kuivila, H . G., Considine, J. L. and Kennedy, J. D. (1972). J . Am. Chem. Soc. 94, 7206 Labinger, J. A. and Osborn, J. A. (1980). lnorg. Cheni. 19, 3230 Labinger, J. A,, Braus, R. J., Dolphin, D. and Osborn, J . A. (1970). J . Chem. Soc. ( D ) 612 Labinger. J. A,, Krdmer, A. V. and Osborn. J . A. (1973). J . Am. Chem. Soc. 95,7908 Labinger, J . A,, Osborn, J. A. and Colville, N. J. (1980). lnorg. Chem. 19, 3236 Landvatter, E. F. and Rauchfuss, T. B. (1982). Organometallics 1, 506 Lappert, M. F. and Lednor, P. W. (1976). Adv. Organornet. Chem. 14, 345 Lau. K. S. Y.. Fries, R. W. and Stille, J. K . (1974). J . Am. Chem. Soc. 96, 4983 Lau, K. S. Y.. Wong. P. K. and Stille, J . K. (1976). J . Am. Chem. Soc. 98, 5832 Louw, W. J., de Waal. D. J. A,, Gerber. T. I. A., Demanet. C. M. and Copperthwaite, R. G . (1982). Inorg. Chem. 21, 1667 Masters, A. F. (1981). “Homogeneous Catalysis, A Gentle Art”, p. 97. Chapman and Hall. London McCleverty, J. and Wilkinson, G . (1964). J . Chem. Soc. 4200 Milstein, D. (1982). Orgunometallics I , 1549 Milstein, D. (1984). Arc. Chem. Res. 17, 221 Milstein, D. and Calabrese, J. C. (1982). J . Am. Chem. Soc. 104. 3773 Moro, A,, Foa, M. and Cassar, L. (1980). J . Organomet. Cham. 185, 79 Mortimer, C. T., Wilkinson, M. P. and Puddephatt, R . J. (1979). J . Organornet. Chem. 165, 265 Mureinik, R. J. Weitzberg, M. and Blum, J . (1979). Inorg. Chem. 18. 915 Nitay, M. and Rosenblum, M. (1977). J . Organomet. Chem. 136, C23 Noyori, R., Umeda, I. and Ishigami, T. (1972). J . Org. Chem. 37, 1542 Ochiai, E. I., Long, K. M., Sperati, C . R. and Busch, D. H. (1969). J. Am. Chem. Soc. 91, 3201 Ohtani, Y.. Fujimoto, M. and Yamagishi. A. (1977). Bull. Chem. Soc. Jap. 50, 1453 Oliver, A. J . and Graham, W. A. (1971). Inorg. Chem. 10. 1165 Pannell. K. H. and Jackson, D. (1976). J . Am. Chem. Soc. 98,4443 Parshall, G . (1977). Catalvsis 1, 334 Pearson, R. G . (1985). Chem. Rev. 85, 41 Pearson. R. G . and Figdore, P. E. (1980). J . Am. Chem. Soc. 102, 1541 Pearson, R. G . and Gregory, C. D. (1976). J . Am. Chem. Soc. 98, 4098 Pearson, R. G . and Muir, W. R. (1970). J . Am. Chem. Soc. 92, 3519 Pearson. R. G. and Poulos, A. T. (1979). lnorg. Chini. Acta 34, 67 Pearson, R. G . and Rajaram, J. (1974). lnorg. Chem. 13, 246 Perland. R. A. and Bergman, R. G . (1984). Organometallics 3, 508 Poli, R., Wilkinson, G., Motevalli, M. and Hursthouse, M. B. (1985). J . Chem. Soc. Dalton Trans. 931 Pribula, C. D. and Brown, T. L. (1974). J . Organomet. Chem. 71. 415 Puddephatt. R. J. and Scott. J. D. (1985). Organometallics 4. 1221

62


Ramasami, T. and Espenson, J. H. (1980). Inorg. Chem. 19, 1846 Saillard, J. Y. and Hoffmann, R. (1984). J . Am. Chem. SOC.106, 2006 San Filippo Jr, J. and Silbermann, J. (1981). J . Am. Chem. SOC.103, 5588 San Filippo Jr, J., Silbermann, J. and Fagan, P. J. (1978). J . Am. Chem. SOC.100, 4834 Schenck, T. G., Milne, C. R. C., Sawyer, J. F. and Bosnich, B. (1985). Inorg. Chem. 24, 2338 Schrauzer, G . N. and Deutsch, E. (1968). J . Am. Chem. SOC.90, 2441 Schrauzer, G . N. and Deutsch, E. (1969). J. Am. Chem. SOC.91, 3341 Schrauzer, G . N., Deutsch, E. and Windgassen, R. J. (1968). J . Am. Chem. SOC.90, 244 1 Sen, L. A. and Halpern, J. (1978). J . Am. Chem. SOC.100, 2915 Shaw, B. L. and Stainbank, R. E. (1972). J . Chem. SOC.Dalton Trans. 223 Shriver, D. F. (1970). Acc. Chem. Res. 3, 231 Sisido, K.. Kozima, S. and Takizawa, K . (1967). Tet. Lett. 33 Slack, D. A . and Baird, M. C . (1974). J . Chem. SOC.Chem. Commun. 701 Slack, D . A. and Baird, M. C. (1976). J . Am. Chem. SOC.98, 5539 Sokolov, V. I. (1976). Inorg. Chim. Acta 18, L9 Steiger, H. and Kelm, H. (1973). J . Phys. Chem. 77, 290 Stille. J. K. and Lau, K. S. Y. (1976). J . Am. Chem. SOC.98, 5841 Stille. J. K. and Lau, K. S. Y. (1977). Acc. Chem. Res. 10, 434 Thorne. D. L. (1980). J . Am. Chem. SOC.102, 7109 Toscano. P. J. and Marzilli, L. G. (1984). Prog. Inorg. Chem. 31, 105 Trost, B. M. and Strege, P. E. (1977). J . Am. Chem. SOC.99, 1649 Trost, B. M. and Verhoeven. T. R. (1976). J . Am. Chem. SOC.98, 630 Trost. B. M. and Verhoeven, T. R. (1978). J . Am. Chem. SOC.100, 3435 Trost, B. M. and Verhoeven. T. R. (1980). J . Am. Chem. SOC.102, 4730 Tsou, T. T. and Kochi. J. K. (1979). J . Am. Chem. SOC.101, 6319 Uguagliati, P., Palazzi, A., Deganello, G . and Belluco, U. (1970). Inorg. Chem. 9, 724 Ugo, R., Pasini, A,, Fusi, A. and Cenini, S. (1972). J . Am. Chem. SOC.94, 7362 Vaska. L. (1968). ACC. Chem. Res. 1 , 335 Van Duong, K. N. and Gandemer, A. (1970). J . Organomet. Chem. 22, 473 Watanabe. Y., Mitsudo, T., Tanaka. M., Yamanioto, K., Okajima, T. and Takegami, Y. (1971). Bull. Chem. SOC.Jap. 44, 2569 Watson, P. L. and Bergman, R. G . (1979). J . Am. Chem. SOC.101, 2055 Wax, J., Stryker, J. M., Buchanan, M., Kovac, C. A. and Bergman, R. G. (1984). J . A m . Chem. SOC.106, 1121 Weber, J. H. and Schrauzer, G. N. (1970). J . Am. Chem. Soc. 92, 726 Whitesides. G . M. and Boschetto, D. J. (1969). J . Am. Chem. SOC.91, 4313 Whitesides, G . M. and Boschetto, D. J. (1971). J . Am. Chem. SOC.93, 1529 Wong, P. K., Lau, K . S. Y. and Stille, J. K. (1974). J . Am. Chem. SOC.96, 5956 Yang, G . K. and Bergman, R. G. (1983). J . Am. Chem. SOC.105, 6045 Yoneda, G . and Blake, D. M. (1981). Inorg. Chem. 20, 67 Zeimer, E. H. K., DeWit, D. G . and Caulton, K. G. (1984). J . Am. Chem. SOC.106, 7006 Ziegler, T. (1985). Organometallics 4, 675

Isotope Effects on nmr Spectra of EquiIibrat ing Systems" HANS-ULLRICH SIEHL Institute of Organic Chemistry, University of Tubingen. D-7400 Tubingen, Germany I

Introduction 63 Equilibrium isotope effects 65 Isotope effects on nmr chemical shifts of static molecules 71 Effect of chemical equilibria on nmr spectra 73 Influence of isotopic perturbation on nmr spectra of equilibrating systems 74 2 Applications 82 Proton tautomeric systems 82 Metalomeric and valence isomeric complexes and carbanions 85 Valence isomerism 91 Conformational equilibria 98 Bridging and hypercoordination in transition-metal complexes 108 Hydrido-bridged carbocations 1 18 Carbon hypercoordinated carbocations 123 Hyperconjugation in carbocations 146 References 158

1

Introduction

The investigation o f isotope effects on chemical reaction rates and equilibria is a well-established tool in physical organic chemistry, and nmr spectroscopy has become a standard technique for the investigation of structure and dynamics of molecules and persistent intermediates. It is the purpose of this chapter to describe a method and its applications to chemical problems

* Dedicated to the memory of the late Victor Gold. He was among the first to use nuclear magnetic resonance for the determination of isotope effects on equilibrating systems.

./

Copwighi 0 IYR7 Acudemrc P r m Inc. London All righrs repprodudon in ony.form reserved

ADVANCES IN PHYSICAL ORGANIC CHEMISTRY VOLUME 23 ISBN 0 12 033523 9

63

64

HANS-ULLRICH SIEHL

which is based upon a combination of equilibrium isotope effects and nmr spectroscopy. The vast majority of applications deal with deuterium isotope effects; only a few '3C-equilibrium isotope effects are discussed to demonstrate the scope of the method. The 'H-nmr spectroscopic determination of a deuterium isotope effect on the fast conformational equilibria of cycloheptatriene (Jensen and Smith, 1964) and the measurement of the deuterium fractionation factor between t-butyl alcohol and water under conditions of slow exchange in the 'H-nmr spectrum (Gold, 1968) are early examples of the application of the technique to equilibrium isotope effects. The effect of isotope perturbation of fast degenerate equilibria on nmr spectra was first applied by Saunders and coworkers (Saunders et af., 1971; Saunders and Vogel, 1971a,b) to study carbocation rearrangements and subsequently was developed into a tool to distinguish rapidly equilibrating molecules from symmetric molecules using 13C spectroscopy (Saunders and Kates, 1977; Saunders et al., 1977a,b). Since then it has been realised that this method has broad general applicability and the area has progressed to such an extent that an overview of the method and some unification of the various applications is desirable. Introductory reviews with selected applications are available (Saunders, 1979; Kalinowski. 1984). The method is presented in reviews concerned with carbocation rearrangements (Saunders et al., 1980; Ahlberg et al., 1983a). Some applications are mentioned in a series on general isotope effects (Forsyth, 1984) and more references can be found in a review of isotope effects on nuclear shielding (Hansen, 1983). The method and some examples have found their way already into an introductory textbook on reactive intermediates (Vogel, 1979) and into recent books on I3C nmr spectroscopy (Kalinowski et al., 1984) and carbocation chemistry (Vogel, 1985). The first part of this chapter gives an introduction to the nmr technique for measuring isotope effects on degenerate equilibria that is most often used, namely, that which is based on chemical shift differences. Sufficient background information is given to follow the discussion of the applications in the second part. The applications are not organised by classes of compounds throughout but for the line of argument and according to general dynamic and structural features. The selection was made primarily to illustrate the different aspects of the subject. It has been considered better to discuss some applications more thoroughly than others, and the chapter is not intended to be a complete survey of the field. Some areas which are not covered are isotope effects on proton and deuterium exchange with solvent, for example, the water-hydronium ion system (Saunders et af., 1984), deuterium isotope effects on acid and base strength (Halevi et al., 1979), on amino acids (Petersen and Led, 1979) and on hydration of cobalt (11) (Saunders and Evilia, 1985). Solvent-dependent isotope effects on equilibria involving hydrogen bonds in carbohydrates and


65

other compounds containing exchangeable hydrogen atoms (Reuben 1984, 1985a,b,c, 1986a,b) are not considered. Isotope effects on complexation equilibria of shift reagents with methyl ethers (DePuy et af., 1976, 1977) and allenic methyl esters (Hansen and Lang, 1980) are also not further discussed. For convenience and internal comparison and consistency, most of the equilibrium reactions involving isotopic molecules are written as proceeding downhill from the initial state (reactant) on the left-hand side to the final state (product) on the right-hand side of the equilibrium. This gives equilibrium constants, K, which are larger than unity. AGO and AH" are accordingly given with a negative sign in the favoured direction of the equilibrium reaction. For most reactions considered in this chapter KH = 1 for the unlabeiled molecules. The equilibrium isotope effect is defined here as KD/KH and, with KH = I , is then equal to KD. This definition is the inverse of common usage; kinetic isotope effects, for example, are usually defined as the rate ratio kH/kD. In the discussion of specific applications, generally only the evidence from isotope effect investigations is presented; this does not always imply that at least some of the conclusions could not have been reached using other evidence. An attempt has been made to give consistent interpretations within the framework of existing isotope effects without doing violence to the Bigeleisen formalism and the Born-Oppenheimer approximation.

E Q U I L I B R I U M ISOTOPE EFFECTS

An equilibrium isotope effect is observed when the equilibrium constant of a chemical reaction is different for isotopomeric compounds. When the bond to an isotopic atom is broken or formed during the course of the reaction the isotope effect is termed primary and when the isotopic bond is neither being broken nor formed the effect is called a secondary one. Secondary isotope effects may be further classified as a-, p-, etc. effects depending on the distance to the reaction centre. The theory of isotope effects is well established and has been presented in detail in the books by Collins and Bowman (1970), Melander (1960). Melander and Saunders (1980) and Willi (1983). Only some general principles of isotope effects on chemical equilibria are presented here mainly to introduce the formulations and parlance of isotope effect theory. A frame of reference is given which is intended to allow the interpretations of the isotope effect studies to be followed with regard to the specific problems described. The Born-Oppenheimer approximation is the cornerstone of theories dealing with the effect of isotopic substitution on molecular properties. This approximation states that electronic and nuclear motion of a molecule can

66

HANS-ULLRICH SlEHL

be separated. The electronic energy of a molecule depends on the nuclear charges of the constituent atoms and on the number of electrons in the system but is independent of the masses of the nuclei. The electronic energy can be evaluated as a function of the fixed position of the nuclei, and the resulting electronic energy surface is the potential surface for the motion of the nuclei. Isotope effects are then nuclear mass effects resulting from motion of different mass nuclei on the same (isotope-dependent) potential energy surface (Wolfsberg, 1969). Model calculations for simple isotopic exchange equilibria taking into account corrections to the Born-Oppenheimer approximation indicate that, although there is no a priori reason to neglect electronic isotope effects (Wolfsberg and Kleinman, 1973, 1974a,b), the predicted magnitude of electronic isotope effects is very small. For equilibrium isotope effects of organic molecules no experimental data are available which have the required accuracy to confirm or deny failures of the Born-Oppenheimer approximation. For some chemical reactant R, which is in equilibrium (1) with some I PH (1) product P, the equilibrium constant K,, neglecting activity coefficients, can be expressed as (2). If R, and P, are the corresponding species containing a RH

KH

(2)

= [PHI/[RHI

specific isotopic substituent, i.e. deuterium, the corresponding equilibrium constant is K D given by (3). The equilibrium isotope effect is then usually KD

=

(3)

[PDI/[RDI

defined by (4), where KHDis really the equilibrium constant for the isotope KHD

=

=

KH/KD

[PHIIRDI/[PDl[RHI

exchange reaction (5). RH

+

P

D

e

-

R

D

+

PH

The equilibrium constant may be expressed in terms of a ratio of complete partition functions Q of the species involved in the equilibrium as in (6). The

reduced partition functions ratios for nuclear motion can then be used to calculate the isotope effect (Halevi, 1963; Wolfsberg, 1972). According to the formalism developed by Bigeleisen and Mayer (1 947) and also by Melander (1960) calculation of isotope effects requires only the


67

knowledge of the frequencies of all the normal modes of vibration of both isotopic forms in the initial and in the final state. The Bigeleisen equation can be written in general form as (7). The MMI (mass, moment of inertia) KHD= A MMI x A EXC x A ZPE

(7)

term is the contribution due to the ratio of ratios of translational and rotational partition functions between the two isotopic species in the initial and in the final state. The EXC or excitation term factor is the effect caused by population of vibrational energy levels above the zero level. The ZPE or zero point energy term accounts for the difference in vibrational zero point energy between the reactants and the products of the isotopic molecules. For molecules also containing heavier atoms such as carbon, the mass and the moment of inertia (MMI term) are scarcely affected by substitution with isotopic hydrogen. Room temperature and below are close enough to absolute zero temperature that most of the C-H or C-D bonds are in their lowest (zero point) vibrational energy level; thus the EXC term makes no significant contribution. For organic molecules and for hydrogen isotopes, where vibrational frequencies and therefore zero point energy are large, it has been found that the ZPE term is the only major contributor to the isotope effect. This is especially valid for secondary isotope effects of hydrogen which are predominantly in this review. Zero point vibrational energy and the corresponding vibrational frequencies and force constants are the focus of all qualitative discussions explaining secondary deuterium isotope effects. Zero point energy effects are generally represented in terms of potential energy diagrams. The potential energy curve for a one dimensional harmonic oscillator with zero point energy levels for stretching vibrations only (Fig. 1 ) is a considerable oversimplification of the real multidimensional energy surface. Other vibrations which are sensitive to isotopic substitution have to be taken into account for a quantitative evaluation of the isotope effect. The potential energy curve for a C--H and a C-D bond for a given bonding situation are identical according to the Born-Oppenheimer approximation. The shape of the bottom of the potential energy curve governs the force constant for the vibrations. A greater force constant (the force constant being the curvature of the potential energy surface) results from a steeper, more curved potential energy well (Fig. la), i.e. from a more closely confined vibrational motion. The restraining forces may originate from the bonds which hold the hydrogen or deuterium to the rest of the molecule or from non-bonding steric repulsive interactions. For a given bond force constant vibrational frequencies and zero point bonds energy Z P E = thv are lower for C--D bonds than for C-H

68

HANS-ULLRICH SlEHL

FIG. 1 One dimensional representation of the multidimensional potential energy well showing zero point energy levels for C-H and C-D vibrations for two different bonding situations, ( a ) high vibrational frequencies and large zero point energy difference, (6) lower vibrational frequencies and smaller zero point energy difference

because of the reciprocal square root mass dependence of vibrational frequencies. In the harmonic oscillator approximation the force constant F for a stretching vibration frequency is related to the vibrational frequency by the Hooke’s law expression (8), where p is the reduced mass for hydrogen

and deuterium respectively. The difference in zero point energy (AZPE) between C-H and C-D increases with increasing vibrational force constant for a given C-H bond. In a chemical reaction where protons and deuterons are allowed to equilibrate among two or more sites having different bonding situations and thus different zero point energy determined by different vibrational force constants, the heavier isotope deuterium tends to accumulate in those positions where it is most closely confined by potential barriers, that is, where force constants, vibrational frequencies and zero point energy are larger. The converse applies for hydrogen which tends to concentrate at those sites in the equilibrium where force constants are smaller (Fig. 2 ) . This can be formulated as a rule of thumb, a first law of isotopic chemistry, as follows. Having the choice, the heavier isotope prefers to make the stiffer bonds. The term “stiffer” binding (Shiner and Hartshorn, 1972) includes restraints to all vibrational modes and is preferred to “tight” which connotes dissociation energies, or to “strong” which might be associated with a short bond and stretching motions only. Recent calculations, for example, have shown that a-secondary isotopic rate effects on reactions leading to carbocations are largely the result of changes in angle-bending force constants, whereas P-secondary isotope effects on such processes arise from changes in bond stretching and angle-bending motions (Hehre et al., 1983).


69

B

A

C

D

FIG. 2 Potential energy curves for an isotopic exchange equilibrium

No isotope effect on a chemical equilibrium is observed when there is no force constant change between initial and final state for a coordinate involving the isotopic atom. For example, the equilibrium constant for the fast 2,3-hydride shift in 2,3-di-trideuteriomethyl-2-butyl cation [ I ] is unity due to symmetry (Saunders and Vogel, 1971b).

For secondary deuterium isotope effect at not too high temperatures the zero point energy difference between initial and final state for an isotope exchange equilibrium is directly related to the standard free energy change and has an exponential effect on the equilibrium constant. The principal possibilities of the temperature dependence of equilibrium isotope effects are discussed in the literature (Melander and Saunders, 1980; Collins and Bowman, 1970). For most practical purposes to evaluate the temperature dependence of secondary isotope effects it has been found that approximate solutions are sufficient. The low temperature approximation (i.e. when the populations of upper vibrational levels are negligible) of the temperature dependence of the equilibrium isotope effect has the form (10) (Stern ef al.. In K = A - BIT

(10)

1968). With A = A S / R and B = A W / R this gives ( 1 I ) . The normal behaviour of the isotope effect is a smooth monotonic decrease with increasing In K

=

-AHo/RT

+ A.Y/R

(11)

temperature. At indefinitely high temperatures the equilibrium isotope effect vanishes and approaches the classical limit of unity.

70

HANS-ULLRICH SIEHL

A hypothetical exchange equilibrium between cyclopropane and ethylene (Fig. 2 ) may serve as an example to demonstrate the qualitative zero point energy analysis of isotope effects. The total reaction A B P C + D is degenerate in the absence of an isotope because A and C and also B and D are identical pairs. The two bonding situation for C-H bonds which are interchanged are represented by two different energy wells I and I1 symbolising a larger and a smaller vibrational force constant for the bond to the cyclopropyl carbon and to the olefinic carbon respectively. The force constant change in the forward reactions A to D (I to 11) and B to C (I1 to I) is opposite to that in the back reactions D to A (I1 to I) and C to B (I to 11) but the total change is the same in the forward and backward direction. If one proton, i.e. the olefinic proton in B, is substituted by deuterium, the potential energy surface with the site-specific force constants is unchanged, but the degeneracy of the equilibrium is lifted because now different total zero point energies apply for both sides of the equilibrium. The total zero point energy of the left-hand side (ZPE of H in well I + ZPE of D in well 11) is larger than on the right-hand side (ZPE of D in well I1 ZPE of H in well I). The zero point energy difference between H and D vibrations on the right side is less than on the left side of the equilibrium as the hydrogen of A (in well I) loses more energy on conversion to D (in well 11) than deuterium is raised in energy on going from well 11 in B to well I in C. A shift of the equilibrium towards the right-hand side leads to a net decrease in vibrational energy. As expected deuterium accumulates in the stiffer bonding position I which has the higher zero point energy whereas the proton prefers position I1 which has the lower force constant and zero point energy. The fractionation factor, which is the equilibrium constant KF = [C][D]/[A][B], was calculated using the Bigeleisen equation as KF = 1.11 at 25°C (Shiner and Hartshorn, 1972). This reaction might serve as a model for a maximum expected a-deuterium isotope effect in which a CH,=CHD group is changed to a cyclopropyl-D group. An experimental example is the equilibrium isotope effect K = 1.109 at 29°C in 1,5-dimethyl-2-deuteriosemibullvalene [2] (Askani et al., 1982)

+

+

H &&D

L ,

PI The investigation of equilibrium isotope effects on chemical reactions is a study of how different masses of nuclei affect motion on the same potential surface. Thus there is a clear and logical difference between the ordinary kind of substituent effect, where the energy surface changes for each


71

substituent, and an isotopic substitution where it does not. In referring to inductive or steric contribution to an isotope effect the origin of chemical isotope effects in vibrational energy difference should be kept in mind. It should also be remembered that, within the validity of the Born-Oppenheimer approximation, isotopic molecules execute their motions on the same potential energy surface to a very high degree of accuracy. Equilibrium isotope effects constitute a subtle tool to study the structure and reaction dynamics of molecules because the isotope hardly changes the object under examination. An observed isotope effect is rationalised by first relating the observed direction and magnitude to certain vibrational force constant changes and these changes can then be related to geometry and molecular structure. ISOTOPE EFFECTS O N NMR CHEMICAL SHIFTS OF STATIC MOLECULES

Isotope effects on vibrational amplitudes are important to rationalise isotope effects on such properties as dipole moment, nmr chemical shift and fine structure in epr spectra (Wolfsberg, 1969). In contrast to thermodynamic isotope effects on chemical equilibria which result from changes in vibrational force constants and zero point energies between the initial and the final state, nmr isotope shifts of static molecules are single-state physical properties, intrinsic to that molecule and its particular set of vibrational force constants. Intrinsic isotope shifts normally do not change much with temperature whereas equilibrium isotope shifts show large temperature dependences. The only conceivable source of intrinsic isotope shifts within the BornOppenheimer approximation lies in the dependence of zero point vibrational motion on isotopic mass (Forsen et al., 1978). All nmr shifts are averages over vibrational motion on the potential surface rather than values for a fixed nuclear configuration (Batiz-Hernandez and Bernheim, 1967). In an isotopically substituted molecule, the average frequencies can in principle be different due to the different amplitudes of zero point motion. The motion then covers a different portion of the same vibrational energy surface and therefore the average shift is changed (Saunders et al., 1984). The majority of deuterium induced isotope shifts of carbon signals are to higher field and decrease rapidly with the number of bonds separating the observed nuclei from the deuterium. Recently an increasing number of upfield and downfield long range deuterium isotope shifts have been reported (Hansen, 1983; Aydin and Gunther, 1981). For one-bond intrinsic isotope shifts, theoretical explanations based on the vibrational origin of isotope effects are available (Jameson, 1977; Jameson and Osten, 1984a,b). The understanding of long range isotope shifts is still scant and the field is

72

HANS-ULLRICH SlEHL

under active investigation. In accord with the vibrational origin of isotope effects it has been suggested that for cyclohexanes and many other compounds efficient coupling between certain vibrations of the intervening bond system plays the major role in determining the magnitude of long range isotope shifts (Gunther et al., 1984). Recently the first isotope effect of tritium on 13Cchemical shifts has been determined (Kresge et al., 1986). For monolabelled acetone the ratio of tritium to deuterium isotope effects for both the upfield a-effect on the methyl group resonance and the downfield P-effect on the carbonyl group resonance are the same, although the isotope effects on the two resonances are believed to be of different origin (i.e. the so-called “inductive” effect for the methyl group and isotopic perturbation of hyperconjugation for the carbonyl group). The measured ratios (1.424 and 1.41) are close to 1.44. This is the value for the exponent relating tritium and deuterium kinetic isotope effects which was derived regarding the isotope effect as a pure zero point vibrational energy effect (Swain et al., 1958). Long range intrinsic isotope shifts often seem to indicate a difference in electron distribution produced by the isotopic substitution as if an electronic substituent effect were responsible. Isotope effects are sometimes discussed not in the framework of general isotope effect theory but in the more familiar “electronic” parlance of physical organic chemistry. Without denying the ultimately vibrational origin of isotope effects, Halevi (1963) has discussed deuterium isotope effects qualitatively in much the same way as chemical substituent effects. Wolfsberg (1969) has pointed out that such an approach has meaning in terms of amplitude isotope effects and has shown (Wolfsberg, 1972) how this can be rationalised within the framework of isotope effect theory. Recently attempts have been made to model the magnitude of intrinsic isotope shifts using molecular orbital calculations (Servis and Domenick, 1986; Forsyth and Yang, 1986; Forsyth et al., 1986). A perturbation induced by artificial shortening of C-H bonds at the site of isotopic substitution is used to simulate the vibrational differences of isotopic isomers. Intrinsic isotope shifts have been discussed in terms of isotopic perturbation of resonance (Saunders and Kates, 1977), through space perturbation of resonance (Forsyth and MacConnell, 1983). perturbation of hyperconjugation (Servis and Shue, 1980; Servis and Domenick, 1985; Gunther and Wesener, 1982; Ernst rt ul., 1983; Forsyth et al., 1984), angular dependence and steric effects (Yashiro et al., 1986; Majerski et al.. 1985). Deuterium induced intrinsic isotope shifts have been correlated empirically with hybridisation (Gunther and Wesener, 1985; Gunther et al., 1985) and in terms of polar electronic influences in the same way as the substituent-induced chemical shift of nonisotopic substituents (Berger and Kunzer, 1985; Berger and Diehl, 1986).

ISOTOPE EFFECTS

ON NMR SPECTRA

73

EFFECT OF C H E M I C A L E Q U I L I B R I A O N N M R S P E C T R A

Dynamic nmr spectroscopy is well suited for the investigation of chemical exchange processes. The method and its application have been extensively treated elsewhere (Jackman and Cotton, 1975; Sandstrom, 1982; Oki. 1985); thus only a brief summary is given in this subsection. In chemical equilibria several molecular states or several molecular species frequently co-exist. This is the case, for example for the individual conformations taken by a given molecule, for molecules being in a tautomeric equilibrium, for carbocations undergoing hydride shifts, etc. When magnetic nuclei experience different magnetic fields because of different chemical environments, a chemical process which allows the nucleus successively to occupy sites of different environment can produce changes in the nmr spectrum even when the system is at equilibrium and no net chemical reaction occurs. This is especially useful for the investigation of degenerate systems, in which the exchange leads to molecules indistinguishable from the original ones. The following changes are generally observed in the nmr spectrum when a reaction interchanges nuclei among different sites. When the process occurs slowly, all lines expected for that particular structure are visible. As the reaction becomes more rapid, as a result of increased temperature, the lines broaden and eventually overlap and coalesce into one or more broad lines. As the rate becomes still faster, these broad lines sharpen and finally one observes again sharp lines but fewer than in the absence of the rate process. The positions of these lines correspond to the weighted averages of the line positions in the static species. The site exchange processes exhibited by the nuclear spins in systems at equilibrium can be investigated by means of band shape analysis. If the frequency difference (A) between the two exchanging peaks is given in Hz the fast limit beyond which increasing rate does not sharpen the averaged lines appreciably is about (A2). Depending on this frequency separation and the temperature range accessible, barriers as low as 3 kcal mol-' have been measured in some cases, as for example in 2,3-dimethyl-2-butyl cation [3] where a large shift difference is present (Saunders and Kates. 1978).

*

(3a]

HANS-ULLRICH SIEHL

74

If the barrier to rearrangement goes to zero, there is no longer a set of equilibrating structures like [3a] and [3b] but a single intermediate structure which might have higher symmetry, for example [4], would then be the minimum energy structure. The dynamics of molecular rearrangements is thus connected to the question of molecular structure. The fundamental question is whether an internal rearrangement is occurring over a barrier on a multiple minimum energy surface or whether a hybrid structure with a single energy minimum is the stable species. Beyond the limit of fast exchange, dynamic nmr spectroscopy cannot generally distinguish between time-averaged symmetry caused by a dynamic process with a very low barrier and time-independent symmetry of a static structure with a single energy minimum. I N F L U E N C E OF I S O T O P I C P E R T U R B A T I O N O N N M R S P E C T R A O F EQUILIBRATING SYSTEMS

Two types of chemical equilibria may be distinguished depending on whether the equilibrium is degenerate or nondegenerate in the absence of the isotope. A number of isotope effects on nondegenerate equilibria have been investigated by nmr spectroscopy (Robinson and Baldry, I977b; Lloyd, 1978; Booth and Everett, 1980b; Reuben, 1984; Forsyth and Pan, 1985). The major part of applications, however, apply to degenerate equilibria where KH = 1 and the measurement of the equilibrium isotope effect reduces to a measurement of K,,. Strictly speaking a degenerate rearrangement, for example the cation equilibrium (12), can be nondegenerate with respect to symmetry from the viewpoint of isotopes in natural abundance. Depending on the reporter nuclei, which might be a I3C isotope used to monitor by nmr spectoscopy the equilibrium isotope effect of another isotope which could be also 13C, the rearrangements of certain isotopomeric molecules containing two' 3C isotopes can be nondegenerate and isotope effects can be measured even without isotopic enrichment. For example, a composite primary-secondary 3C equilibrium isotope effect has been determined in 2,3-dimethyl-2-butylcation [5] from the shift difference observed in single- and double-quantum 13Cnmr spectra at natural abundance (Olah et al., 1985a). The sensitivity of this experiment is of course severely limited by the low natural abundance of H3C\+ Y m 3

,c-c

H3C

L

*$H3

HC ,5 ,-1

,c-c

H3C

+

,CH3

*iCH3


75

[()I the vicinal 13C-13Cisotopomers [5]. The primary equilibrium isotope effect on (14), essential to extract the pure secondary effect in ( 1 3) is not accessible by this approach but requires specific labelling (Saunders et al., 1981). If the rate of a dynamic process is slow relative to the nmr time scale, separate signals for the isotopic isomers can be observed and the equilibrium constant, i.e. the thermodynamic deuterium isotope effect, can be obtained directly from the value of the ratio of integrals at equilibrium. As integration is normally less precise than frequency measurements, the equilibrium constant can be obtained only with lower precision from area methods than from chemical shift methods (Saunders ef al., 1980~;Booth and Everett, 1980b). Recently, however, it has been demonstrated that the area method is also capable of measuring relatively small equilibrium isotope effects from 'H spectra with high accuracy using the newest generation of very high field spectrometers under very carefully controlled conditions (Rappoport et al., 1985). The chemical shift method for the determination of equilibrium isotope effects is based on the chemical shift difference (the equilibrium isotope shift) between nmr signals for nuclei which are time-averaged to equivalence in the absence of the isotopic perturbation (Saunders, 1979). A rapidly equilibrating system, for example, a two-fold degenerate rearrangement in carbocation [7] shows an nmr spectrum with averaged resonances for those atoms which are interchanged by the dynamic process (15).

[7a1

[7b1

The signal for carbon-I in [7a] and [7b], which has an average shift of the two positions in [7a] and [7b], and the corresponding averaged signal for carbon-2 in [7a] and [7b] appear at the same position because the concentrations of [7a] and [7b] are equal ( K = 1) in the nondeuteriated cation. Unsymmetrical introduction of deuterium breaks the symmetry of the twofold degenerate rearrangement ( 1 6 ) . The equilibrium is perturbed if there are zero point energy differences resulting from differences in vibrational force constants for the isotopic isomers [8a] and [8b].

v',,,)

HANS-ULLRICH SIEHL

76

A

c-2

Ill\

J

c-l

/

\ \

\

/

/

A+B

B

C-1

The change in free energy which is related to the zero point energy difference will be different from zero and with AGO = -RTlnK and K = [B]/[A] where [A] and [B] are the concentrations of [8a] and [8b] the equilibrium constant K is not equal to unity. The resulting concentration difference between [8a] and [8b] leads to a population difference of the exchanging sites for C-l and C-2 in (16) and this lifts the chemical shift degeneracy of the averaged C- 1 and C-2-signals which was present in (1 5). The chemical shift for C-l is the concentration-weighted averagef, of the two shifts F , = F,+ and F2 = FCHof the two sites in [8a] and [8b] and is given by (1 7). I f the equilibrium is shifted to the right preferring [8b], site A is

less populated than site B. The averaged signal for C-l is observed off'set the shift is from the unperturbed averaged position faYe, = i(F,+ + FCH);

ISOTOPE EFFECTS

ON

NMR SPECTRA

77

downfield towards the chemical shift F, = F,, . Analogously, the averaged signalf, of C-2 in [8a] and [8b], given in (18), is offset from the unperturbed f 2

= (Fi[Al

+ Fz[Bl)/([Al + P I )

(18)

position in the opposite direction, shifted upfield towards the chemical shift F, = F C H of C-2 in [8b]. The difference 6 = f , - f , is the equilibrium isotope shift difference. The ratio of the isotope shifts for the peak in the downfield position (f,- f,,,,)and in the upfield position (fa,,, - f , ) of the peaks of [8] at positionsf, and f , relative to the peak of [7] atfa,,, is 1 : 1, because a two site fast exchange is taking place between a singly populated highfield and a singly populated low field site. Depending on the distance of the deuterium substitution from the other exchange site appropriate corrections for the deuterium-induced intrinsic chemical shift (&,,,,) must be taken into account to evaluate the splitting 6 (6 = ?jobs- 6intr)caused by the chemical isotope effect. The size of the splitting 6 between the averaged lines is related to the chemical shift difference (A = F, - F, = F,, - FCH)in the spectrum of the cation at the slow exchange limit and to the isotopic equilibrium constant K = [B]/[A]. The equation relating the isotope splitting 6 to the equilibrium constant K may be derived as follows. From (1 7) and ( 1 8), 6 is given by (19),

and substituting K

=

[B]/[A] one obtains (20). This equation can be simpli6

fied using A

=

F,

=

F,K - FZK - Fl I f K

+ F2

F, yielding (21) and (22). Equation (22) relates the 6=--

KA - A I+K

K = - -A + 6 A-6

observed equilibrium isotope effect shift difference 6 to the isotopic equilibrium constant K in a two site fast exchange equilibrium where two nuclei are interchanged between two singly populated sites. An analogous calculation for the methyl groups or the averaged methine carbons in the triply degenerate rearrangement (23) of 1-deuterio-l-(p,Pdimethylcyclopropy1)ethyl cation [9] which is a two site fast exchange

7a

HANS-ULLRICH SIEHL

HI H

interchanging three nuclei between a singly populated (low field) site and a doubly populated (high field) site yields (24). K = -A f S A - 26

If the fast dynamic process can be frozen out on the nmr time scale by lowering the temperature, A can be determined directly. Otherwise it must be estimated using suitable reference data from compounds which might serve as a model. There are several possible sources of error in the determination of K . One source of error is A which might not be accurate either because the lines under slow exchange conditions are still broad or the slow exchange area is not accessible and the estimate for A from model compounds may not be accurate enough. The error in K then depends on the relative ratio of A and 6. Large values of A and small values of 6 give the smallest error in K . It is often advantageous to use a reporter nucleus which has a large shift range such as 13C, "F, etc. Also the precision may be increased by using high field spectrometers. As the isotope splitting 6 normally increases with the number of deuterium atoms at equivalent positions in the molecule, the error in K increases for multi-deuteriated compounds. The accurate determination of 6 can be a source of error when the &-valueswhich are dependent on the size of A and on the equilibrium constant K are small. High field measurements can increase the accuracy but when the lines are kinetically broadened the accessible temperature range for accurate determination of 6 may be smaller.


79

The corrections necessary to consider intrinsic isotope shifts may be another source of error, especially in those cases when 6 is small. Sometimes several different signals can be observed which are averaged by the same dynamic process. The same equilibrium constant K applies then for all isotope splittings observed, but each set of averaged signals has a different A and thus different 6. The size of 6 depends on the size of A and the equilibrium isotope effect K which measures the difference between local force fields at the exchanging sites and can be very large. The 1-trideuterio-2trideuteriomethyl-3-methyl-2-butylcation [lo], for example, shows an isotope splitting of 6 = 167.7 ppm for the averaged C+/C-H carbons in the 13C spectrum at - 138°C (Kates, 1978).

The isotope splitting 6 varies strongly with temperature and this is an important diagnostic characteristic that an equilibrium isotope effect is being observed. From the temperature dependence of 6 the temperature dependence of K and AGO = - R T In K can be determined. If In K is found to vary linearly with 1/T in the temperature region studied AH" and A F can be ( A F / R ) by determined from a van't Hoff plot of In K = -AH"/RT regression analysis. For systematic errors in the determination of thermodynamic parameters for equilibrium isotope effects using these chemical shift methods, criteria apply similar to those discussed in the literature on the determination of activation parameters from dynamic nmr measurements. Precision and accuracy are dependent on the temperature range studied and on the accuracy of the temperature measurement. AS" exhibits a strong tendency to become zero with improvement of the accuracy of the experiment. In special cases an inverse temperature dependence of an equilibrium isotope effect increasing at higher temperatures has been observed (Kates, 1978; Walter, 1985). When an equilibrating system, that exhibits a small isotope effect, is separated by only a small energy barrier from a structurally different system and this interchange causes a larger isotope effect, the observed effect will increase at higher temperatures. The kind of information obtainable through the use of the isotopic perturbation method is manifold and the problems which can be investigated cover a broad range. Fundamental questions connected to the interplay between molecular dynamics and structure like the distinction between time-averaged symmetry of molecules with a multiple minimum energy

+

80

HANS-ULLRICH SIEHL

surface and molecules having a single energy minimum can be made using this technique. The occurrence of sizeable temperature-dependent isotope splittings resulting from a thermodynamic isotope effect is characteristic of a dynamic equilibrium. In molecules which have a single minimum energy surface, for example, a resonance system like I-deuteriocyclohexenyl cation [ l I], there is no equilibrium and hence no thermodynamic isotope effect is observed (Saunders and Kates, 1977). The intrinsic isotope shifts in single state structures are much smaller and not very sensitive to changes in temperature.

As long as the activation energy for equilibration in a multiple energy minimum system is not below the zero point vibrational energy difference of the isotopomeric molecules, the height of the energy barrier has no direct bearing on the strength of an equilibrium isotope effect. The nmr test for isotope perturbation of degenerate equilibria is still decisive even when no kinetic line-broadening is observable in systems rearranging over a very low barrier. Tautomerism, valence-isomerism, bridging and hypercoordination' have structural implications where the question of whether an observed atom equivalence is time-averaged or not is relevant, and representative examples have been submitted to the perturbation of degeneracy test. The method is not limited to the purpose of demonstrating the existence or the absence of an isotope effect. Provided that the chemical shift difference A is accessible. it is a very accurate method for the determination of different kinds of equilibrium isotope effects. The isotope effect is calculated from nmr frequency differences which can be measured with high precision. The method does not depend on knowing the exact amount of isotope incorporation in starting materials or products, as in other methods for determining equilibrium isotope effects. Its usefulness is limited however to those molecules undergoing degenerate rearrangement rapidly enough to give averaged spectra. The species investigated are stable molecules or persistent intermediates which are structurally well defined by nmr spectroscopy. Data obtained by

' A carbon or hydrogen atom is hypercoordinated if the number of attached atoms exceeds four or one respectively.


81

this method may thus be useful as an experimental test of isotope effect theory and calculation procedures and can be used to determine isotopic fractionation factors, for example for molecule-solvent hydrogen exchange (Saunders and Jarret, 1985, 1986). Equilibrium isotope effects serve as an empirical calibration of kinetic isotope effects. P-Secondary deuterium isotope effects in equilibrating carbocations have been used especially to prove the interpretation of the hyperconjugative origin and the dihedral angle dependence of P-kinetic deuterium isotope effects in nucleophilic substitution. Besides the answer to fundamental questions and the test of general principles this method can give access to molecular structural details which are not easily accessible otherwise. The applicability of the isotope method to study conformational equilibria rests on the fact that the increase in force constant and frequency for all isotopes of hydrogen at sterically more constrained positions increases the separation between zero point energy levels of isotopic oscillators. The isotopic perturbation method is very useful for studying fractional bonding and hypercoordination in coordinatively unsaturated and electron deficient compounds such as transition metal complexes or carbocations. The 2-norbornyl cation and the bicyclobutonium cation are the most prominent examples of carbocations whose structures have led to much controversial discussion (the so-called classical-nonclassical ion controversy). The application of the isotopic perturbation methed is likely to be the most decisive piece of nmr evidence for the hypercoordinated structure of these two cations in solution. In this context it is very useful that this technique provides some general evidence on exchange characteristics which can be evaluated from the spectra of the deuteriated isomers independent of whether the slow exchange region is accessible or not. Distinction between equilibria which are degenerate or not in the absence of the isotope can often be made from the temperature dependence of chemical shifts of the unlabelled molecule but also from the fact that no isotopic perturbation is observed for nuclei positioned on a symmetry element common to all averaged deuteriated isomers. The isotopic splitting patterns allow the determination of the number of exchanging sites and their relative population. In most cases the direction of an isotope effect can be extracted from 'H or 13Cnmr spectra even though quantitative evaluation of the isotope effect is not always possible. The direction of the isotope effect gives important information on the relative stiffness of vibrational bond force constants which in turn can be related to characteristic features of the alternative structures under consideration.

HANS-ULLRICH SlEHL

a2

2 Applications P R O T O N TAUTOMERIC SYSTEMS

2,4-Pentanedione Different possibilities for the structure and dynamics of the cis-enol of 2,4-pentanedione have been suggested: the symmetrical structure [ 131 (Shapet'ko et al., 1975; Karle et af., 1971), an equilibrium between two structures [12a] and [12b] (Robinson et al., 1977; Forsen et al., 1978) or an equilibrium between all three structures (Chan et af., 1970; Leipert, 1977). In the case of a potential energy surface with shallow minima, deuteriation of -OH could lead to large temperature-dependent intrinsic isotope shifts caused by an amplitude isotope effect on the vibrational frequencies. Therefore the test for isotopic perturbation of symmetry is a more conclusive distinction between a single- and a double-well potential energy surface.

[ I2a]

[12b]

The 13Cnmr spectrum of the enol tautomer of 2,4-pentanedione in CDCl, shows a single time averaged peak for the carbonyl carbons. Unsymmetrical introduction of deuterium was used by Saunders and Handler (1 985) to break the symmetry. In a mixture of 1-mono-, di- and trideuterio-2,4-pentanedione [14] (x = 1, 2, 3) a pair of peaks was observed for the carbonyl carbons in each isotopic isomer. The splitting at 23°C was 0.167 ppm per deuterium and was found to be temperature-dependent, decreasing with increasing temperature. This is characteristic of an equilibrium isotope effect and indicates a double minimum energy surface and an equilibrium (26) between [ 14a] and [ 14bI. Analysis of the temperature dependence yielded A H o = - 14.4 cal mol-' per D.

83


Conclusive proof comes from an investigation of the unsymmetrically deuteriated bis-2,4-pentanedione complex of zinc. The carbonyl resonance of this complex is a single peak and no splittings were observed. This shows that the anion of the zinc complex is a symmetrical single energy minimum and confirms the conclusion that the enol is a rapidly equilibrating mixture of [12a] and [12b]. The direction of the equilibrium isotope effect in the enol was determined from the observation, that for [14] ( x = 1, 2, 3) the downfield signal of each isotopic peak pair was a singlet and the upfield signal showed deuterium couplings. The singlet signal results from the averaged carbons which are further from the deuterium. The signals found shifted downfield towards the ketone position indicate that the isomer [14b] with the deuteriated methyl group at the enol carbon is preferred in the equilibrium. Thio1-hydroxy and amino-hydroxy tautomerism Hansen (1 983) has reviewed a number of primary deuterium isotope effects in other tautomeric systems and only some recent examples are given here. Deuterium isotope shifts over at least six bonds were observed in rapidly interconverting p-thioxoketone tautomers [ 151 using 3C-nmr spectroscopy (Hansen et al., 1982). The observed effects are caused by a shift of the fast enol-enethiol equilibrium (27) when the chelating proton is substituted by deuterium. Deuterium prefers attachment to oxygen as compared to sulphur

'

~ 5 1

because oxygen-hydrogen(deuterium) bonds are shorter and thus stiffer than thiol S-H(D) bonds. The magnitude (up to 8 ppm) and the sign of the isotope shift depend on the chemical shift differences of the exchanging sites and on KH for the nondegenerate equilibria. P-Thioxo-esters and thiolesters

a4

HANS-ULLRICH SlEHL

show only small deuterium isotope effects (Hansen and Duus, 1984); in p-thioacids the tautomerism does not take place and only intrinsic isotope effects are observed. Deuterium isotope effects in nondegenerate tautomeric equilibria where the chelating H(D) is bonded either to an oxygen or a nitrogen atom have been investigated in enaminoketones [I61 (Shapiro ef al., 1981) and Schiffs bases of salicylaldehydes [ 171 where the nonaromatic tautomer is strongly disfavoured in equilibrium (29) (Hansen er al., 1982). An intrinsic origin for the small effects ( k 0.4 ppm) in enaminoketones [I 61 was envisaged by Shapiro, whereas Hansen suggested an equilibrium perturbation effect or a combination of both influences.

Deuterium isotope shifts have been observed in the a-ketohydrazone/ azoenolic tautomeric compound [I81 by 13C and "N nmr spectroscopy (Hansen and LyEka, 1984). Besides large intrinsic isotope shifts, long range equilibrium isotope effects were observed which are caused by perturbation of the tautomeric equilibrium (30). The shifts increased on lowering the

temperature as expected for equilibrium isotope effects. Deuterium substitution shifted the equilibrium in all cases further in the direction of the tautomer which is predominant in the protio compound.


85

METALOMERIC A N D VALENCE ISOMERIC COMPLEXES AND CARBANIONS

Cyclopen tadienyl complexes Rapid migration of the metal around the ring in many q'-cyclopentadienyl complexes results in averaged carbon and proton signals for the ring atoms. If the barrier of this process is low, no line-broadening can be observed even at very low temperatures. In this case nmr spectroscopy cannot distinguish between pentahapto coordination and a fast equilibrium of monohaptocoordinated species. Two key systems [I91 and [20] for which the structures are known were investigated by Saunders et al. (1980b) to show that a distinction can be made between a fluxional q '-cyclopentadienyl ring and a q5-cyclopentadienyl ring using the isotopic perturbation technique. The proton coupled 3C spectrum of a mixture of 1 ,l'-bisdeuterioferrocene and nonlabelled compound showed a triplet for the deuteriated carbons shifted 0.178 ppm upfield from the centre of the doublet (JCH = 177Hz) which was assigned to the methine carbons in the deuteriated and unlabelled ferrocene. The shift is found to be temperature-independent as expected for a normal intrinsic isotope shift in a static system and is thus consistent with symmetrically bound pentahapto coordination and a single minimum structure of ferrocene [19].

I

The spectrum of a mixture of monodeuterio and perprotiocyclopentadienyltrimethyltin on thc contrary showed for the triplet of the deuteriated carbons a downfield shift of 0.324 ppm at 25°C relative to the centre of the methine carbon doublet from the nondeuteriated ring. More accurate isotope shifts and better resolved signals for the p- and y-carbons were obtained from proton-decoupled I3C nmr spectra. Significant increases in the isotope shifts were observed when the temperature was lowered, indicating a definite equilibrium isotope effect in cyclopentadienyltrimethyltin.The tin complex is an equilibrium mixture of three different isotopic isomers [21a-c]. The fast exchange takes place between isomer [2 la] with deuterium at the metal-substituted a-carbon and isomers [21b] and [ ~ I c with ] deuterium at the p- or y-olefinic carbons (31). A peak for the single a-carbon showed deuterium coupling and was found shifted downfield, whereas the

HANS-ULLRICH SlEHL

86

M

pM

q [21b]

(31)

O

0

6

D M = Sn(CH313

peaks for the two p- and two y-carbons are shifted upfield relative to the shift of the unlabelled complex [20]. The chemical shifts for the metal-substituted a-carbon and the p- and y-olefinic carbons in a static cyclopentadienyl tin complex can be estimated by considering the chemical shifts in limiting slow exchange spectra of suitable model compounds. The metal-substituted carbon in [22], [23] and [24] have chemical shifts of 30-50ppm and the olefinic carbons appear at 120-150 ppm. Comparison with the observed isotope shifts in the tin complex [21] shows that the deuterium prefers the olefinic carbon relative to the metal-substituted carbon. The isomers [21b] and [21c] with deuterium at the olefinic position are therefore preferred in the equilibrium.

[22]

~ 3 1

P41

The direction of this equilibrium isotope effect is the same as in the allylmagnesium bromide [26] (see below) and is the reverse of what is observed in isotopic exchange equilibria of unsubstituted hydrocarbons with similar structural differences. In the Cope rearrangement (32) of the

ISOTOPE EFFECTS

ON NMR SPECTRA

a7

[25a]

tetradeuteriated- 1,5-hexadiene [25], deuteriums bonded to terminal sp2-hybridised carbons are interchanged with deuteriums bonded to sp3-hybridised carbons and preferred at the sp3-hybridised carbon site in [25b] (Sunko et al., 1970). Since deuterium isotope effects are determined in large measure not only by the number but also by the kind of atoms directly attached to the carbon with the exchanging hydrogen or deuterium, it might be anticipated that the valence bending motions of C-H and C-D bonds which are predominantly responsible for a-secondary isotope effects are very different at the carbon bonded to a metal in [20] as compared to an sp3-hybridised aliphatic carbon in [25]. If it is assumed that there is no force constant difference and thus no equilibrium isotope effect between the P- and y-olefinic positions on [21], the different upfield shift observed for these carbons is due to different chemical shifts for the two positions in the “frozen” static spectrum. The isotopic exchange would then occur between a single populated downfield site, the a-position, having different vibrational force constants from the upfield side which has four times the population (the p,p’ and y,y‘-positions). The spectra indicate that the sum of the upfield shift of the two olefinic positions is half the size of the downfield shift of the deuteriated a-carbon when corrections for intrinsic shifts are taken into account. The observed shift difference between the deuteriated carbon and the P-carbon was corrected for intrinsic isotope shifts which were taken from model compounds. This gave an equilibrium isotope splitting of 6 = 0.589 ppm at 25°C. The approximate formula for calculating the equilibrium constant is K = (A + S)/(A - 46) assuming equivalent olefinic positions. Using A = 90ppm as an approximate shift difference between the metal-substituted and olefinic carbons. one obtains K = 1.034 at 25°C. Ally1 derivatives of the alkali metals and alkaline earth metals Dynamic nmr spectroscopy even at very low temperatures shows apparent symmetry for the allylmetal compounds [26] indicating either a very low barrier for the metalomeric equilibrium (33) between degenerate monohapto species [26a] and [26b] or the presence of a n-bridged species [26c] with time-independent symmetry. Deuterium isotope effects on terminal monoand geminal-dideuteriated ally1 compounds of the alkali and alkaline earth metals were investigated to allow distinction between these two possibilities

HANS-ULLRICH SIEHL

88

M’

H (33)

H -

H [7ha]

H [Xh]

1

M =MgBr

M2=Li, No, K.Rb. Cs

(Schlosser and Strahle, 1980, 1981; Schleyer and Neugebauer, 1980; Bywater et al., 1980). In monodeuteriated allylmagnesium bromide the averaged terminal carbons give two separate signals, a triplet for the deuteriated carbon shifted downfield and a singlet for the nondeuteriated carbon shifted upfield with respect to the peak of the unlabelled compound. The splitting is about 1.9 ppm at room temperature when an intrinsic upfield shift of 0.4 ppm for the deuteriated carbon is taken into account. At lower temperatures the splitting increases as expected for an equilibrium isotope effect. The alkali metal ally1 compounds show splittings that are one order of magnitude smaller. No temperature-dependence was reported or is evident from comparison of different investigations. It was concluded that the allyl magnesium bromide has a o-covalent unsymmetrical structure with a monohapto ligand-metal interaction. A fast, reversible equilibrium (33) interconverts structure [26a] and its tautomeric structure [26b]. In contrast the alkali metal ally1 compounds can be described as more or less symmetric rc-complexes [26c] with trihapto interaction between metal and ligand. The equilibrium constant in terminal monodeuteriated allylmagnesium bromide was estimated to be K = 1.04-1.08 at 24°C using a value of 70 ppm for the shift difference between the olefinic (estimated 92ppm) and the metal-bound (estimated 22 ppm) carbons. The direction of the isotope effect is evident from the downfield shift of the deuteriated carbon towards the olefir.ic position. The equilibrium is shifted so that the isotopic isomer with deuterium at the terminal olefinic position is preferred in the equilibrium. The values for the small isotope shifts and splittings in the deuteriated alkali metal allyl compounds vary slightly in different investigations, most probably due to different experimental conditions i.e. concentration and temperature. In a 90MHz 13C nmr reinvestigation of alkali metal allyl compounds, Schlosser and Strahle (1981) report small upfield shifts for both


89

the deuteriated and the nondeuteriated terminal carbon signals with respect to the peak in the corresponding unlabelled compounds. In the d,-lithium compound, the deuteriated carbon triplet is found downfield, whereas in the potassium and caesium compounds this peak is upfield relative to the nondeuteriated carbons. For the lithium compound it is suggested that bonds of unequal strength bridging the two terminal carbons of the allyl unit would give an unsymmetrical distorted n-complex structure. Very small distortions in the potassium and caesium n-complexes which might even be absent in the unlabelled species were considered to explain these observations. Although the effects observed are small, it is reasonable to envisage a gradual change going from monohapto coordination for magnesium to unsymmetrical bridging for lithium and symmetrical bridging for the higher alkali metals. The situation is similar to that with carbocations where it has been shown that there is no dichotomous contrast between fast rearranging and bridged species, i.e. between a multiple minimum energy surface and a single minimum. Because of the shallow potential wells of these species, small changes in energy can change the structure, and it is best to think of the energy surface as showing a continuous variation.

Bicyclo [3.2.l]octu-3,6-dien-2-yl union The bicyclo[3.2.I]octa-3,6-dien-2-yl anion [27] was proposed as the prototype of bishomoaromatic anions. The isotopic perturbation method has been applied to this system to obtain experimental evidence on its nature (Christ1 et ul., 1983).

Two interpretations have been advanced to explain the upfield shift observed for the C-6,7 carbons (91 ppm) in [27] compared to the corresponding shifts of 130.36 and 139.82 ppm (C-6;C-7) in the parent hydrocarbon [28]. If the anion is a bishomoaromatic static molecule [29], delocalisation of charge from the allyl anion moiety to C-6,7 could account for the observed upfield shift. A multiple minimum energy surface with a rapidly equilibrating mixture of allyl anion [30] and the tricyclic and tetracyclic isomers [31] and [32] would give C-6,7 in part carbanion and part cyclopropyl-carbon character and could also explain the observed upfield shift.

90

HANS-ULLRICH SIEHL

If the anion were equilibrating the deuteriated anions [33] and [34] would be expected to show equilibrium isotope effects since the C-H(D) vibrational bond force constants are different in the valence isomers [30]-[32]. The structures [3I] and [32] containing cyclopropyl rings should be favoured in the equilibrium because C-H(D) vibrations are more confined in a cyclopropane position than at an sp2-hybridised carbon. Only small upfield shifts were observed in the deuteriated anions [33] and [34]. The maximum shift was 0.49 ppm for the deuteriated carbons C-2,4 in the dideuteriated anion [34]. The small magnitude of the shifts and the typical decrease of the long range effects characterise these as being intrinsic rather than resulting from an equilibrium perturbation. The single energy minimum bishomoaromatic structure [29] is in accord with this result. The intrinsic upfield shift at C-6,7 in [33] and [34] is opposite to the downfield shift observed in the 2,4,4-trideuteriated hydrocarbon [28]. This has been taken as evidence for a direct bonding interaction between the ally1 anion terminus and the ethylenic unit. Titanacyclobutanes The important species in the fundamental step of olefin metathesis and related catalytic reactions have been described as either rapidly equilibrating metal-alkylidene complexes or metallacyclobutanes. X-ray crystallographic

ISOTOPE EFFECTS

ON NMR SPECTRA

91

studies on several substituted dicyclopentadienyltitanium complexes have shown that these complexes have a near planar titanacyclobutane structure in the solid state. Deuterium isotope effects have been investigated by Grubbs et al. (1981) in order to establish the structure in solution.

(35)

(361

The "C nmr spectra of C-I-dideuteriated [35] showed two peaks for the C-l and C-2-methylene groups. The quintet of the deuteriated methylene C-l was shifted upfield by 0.7914.758 ppm compared to the corresponding protio compound, whereas the nondeuteriated methylene group C-2 was shifted downfield by 0.121, 0.060 and 0.045ppm depending on the C-3 substituent. All other signals showed intrinsic isotope shifts upfield. The temperature-independence of the shifts and the small ratio of the observed isotope splitting to the estimated shift difference (100 ppm) between the averaged C- 1 /C-2 carbons in a titanium-alkylidene-olefin structure [36] were interpreted as being consistent with a symmetrical but easily distorted titanacyclobutane structure [35] resting at the minimum of a broad shallow potential energy surface which allows easy distortion towards a transition state for the metathesis reaction. V A L E N C E ISOMERISM

Barharalone At room temperature barbaralone undergoes a rapid Cope rearrangement (36) between two identical structures [37a and b] which leads to averaged signals for H- 1 /H-5 and for H-2,H-8/H-4,H-6.

92

HANS-ULLRICH SIEHL

In 1 -deuteriobarbaralone [38], deuterium lifts the degeneracy of the equilibrium (37) and two peaks for the averaged H-2,H-8/H-4,H-6 protons were observed separated by 0.168 ppm at room temperature (Schleyer et al., 1971). From the limiting, slow exchange spectrum of [37] an intrinsic shift difference of 3.04 ppm between the H-2,H-8 and H-4,H-6 protons was determined. Integration of the signals for the cyclopropyl proton at 2.46 ppm provided the direction of the equilibrium isotope effect and showed that deuteriums are preferentially attached at the bridgehead position in [38b]. The equilibrium constant can be calculated using the formula K = (A + &)/(A- 6); with A = 3.04ppm and 6 = 0.168ppm, K = 1.117 at room temperature, corresponding to a ratio of 53 : 47 for [38a] : [38b] and AGO = -65 cal mol- ' at 25°C. 0

L J QD

0

J

7

Q D

(37)

\

[38a]

[38b]

The equilibrium constant recalculated from the shift values reported is in reasonable agreement with the fractionation factor for a simple model reaction of isotopic exchange equilibrium (38). The fractionation factor, Kf = 1.07 (25"C), for d,-cyclopropane relative to 2-d1-propane [39] (Shiner and Hartshorn, 1972) might serve as a model value for the maximum expected a-deuterium isotope effect.

D H

A

H H +

CH3CH,CH3

CH3CHDCH3 + ~ 9 1

21 (38)

The equilibrium isotope effect in [38] is a differential a- vs y-effect. In 1-d,-semibullvalene [45] a comparable exchange situation exists for deuter-


93

ium bonded either to a cyclopropyl carbon or to a bridgehead aliphatic carbon and similar isotope effect data have been reported by Kalinowski (1984). The degenerate Cope rearrangement of bicyclo[5.1 .O]octa-2,5-diene (3,4-homotropylidene) which is a non-bridged analogue of [37] is perturbed by deuterium. Analysis of 100 MHz 'H nmr spectra of an unsymmetrically octadeuteriated isomer revealed that CD, is favoured in the diallyl position and CH, in the cyclopropyl position in accord with the results obtained for [38] and [45] (Giinther et a / . , 1975).

0

0

Although quantitative evaluation of the equilibrium isotope effect in 2-deuteriobarbaralone [40] was not possible, the direction of the isotope effect favouring deuterium bonded to the cyclopropyl carbon in [40b] over the olefinic position in [40a] could be established. The sign of the equilibrium isotope effect is in agreement with what would be predicted from force constant calculations of model exchange reactions and also with the equilibrium isotope effect observed in 1,5-dimethyl-2-deuteriosemibullvalene[46] (Askani et a / . , 1982). 9-Barharalyl cation The 9-barbaralyl cation (C,H,+) can be regarded as cationic analogue of barbaralone. This cation has been reviewed by Ahlberg et a / . (1983a) and only a brief summary of the decisive experiments is given here. The 13C nmr spectrum of C,H,+ at -135°C shows only one broad peak indicating a dynamic process which averages all nine carbons. At - 15 I "C this peak is split into two peaks at 101 ppm and 152 ppm with an intensity ratio of 6 : 3. [42] both peaks show large isotope In the octadeuteriated cation C,D,H shifts as compared to the unlabelled ion, indicating that the spectrum at - 151°C does not result from a static symmetrical cation but that both peaks are still averaged by a fast equilibration process which is perturbed by deuteriums (Ahlberg et al., 1981). The lowest energy cation [41] has thus a structure which undergoes a partial (six-fold) degenerate rearrangement, and the averaging process with the lowest barrier is not a Cope-type rearrangement but a divinylcyclopropylmethyl-divinylcyclopropylmethylcation rearrangement (40). +

HANS-ULLRICH SlEHL

94

JI

In the C,D,H+-cation [42] a downfield shift of 6 ppm is observed for the six averaged carbons which appear at 101 ppm in C,H,+. The three averaged carbons at 152ppm in C,H,+ are observed 1 ppm upfield in C,D,H . These equilibrium isotope shifts can be rationalised by taking into account the differences of the C-H bond vibrational force constants of the exchanging carbons (Ahlberg et al., 1983b). The upfield signal results from the averaging of six carbons (C-1, C-5, C-2, C-4, C-8, C-6), two of them olefinic and the other four saturated C-H bonds. Since vibrational force constants are lower in olefinic C-H bonds than saturated C-H bonds, C-H will be preferred in the former and C-D in the latter positions. By comparison with the I)-CH,-substituted cation, the olefinic carbons are +


95

H

expected to absorb at lower field than the aliphatic carbons in a static 9-barbaralyl cation. Therefore a downfield shift for the C-H signals of the six averaged carbons in the C,D,+-ion is observed. The smaller upfield shift (1 ppm) for the three averaged carbons, C-9, C-3 and C-7 (one charged carbon and two olefinic carbons) at 152 ppm has been suggested to result from a slight preference for the C-H groups to occupy the olefinic positions. At - 135°C the totally degenerate rearrangement having a higher barrier occurs via [43] as an intermediate or transition state (Ahlberg et al., 1983c) and this averages the two equilibrium isotope effects, resulting in a downfield isotope shift of 4.5 ppm.

V31

Semibullvalene Semibullvalene [44] has a very low barrier to Cope rearrangement (4 1) which averages C-l/C-5 and C-2,C-S/C-4,C-6 and leaves C-3 and C-7 unaffected. Below - 160°C the exchange is slow on the nmr time scale and the averaged carbon and corresponding protons signals are split, giving five resonances as expected. The carbon shift differences are 5.8 ppm and 89.6 ppm for C-I/C-5 and C-2,C-S/C-4,C-6 respectively. The cyclopropyl carbons C- 1 and C-2,C-8 are shifted upfield (Anet et al., 1974).

96

HANS-ULLRICH SIEHL

Equilibrium isotope effects in deuteriated derivatives of [44] were investigated by I3C nmr spectroscopy by Askani et al. (1982, 1984) and have been reviewed in detail (Kalinowski, 1984), As expected, the largest isotope splittings are observed for the averaged C-2,C-8/C-4,C-6 carbons which have the largest shift difference at slow exchange. The direction of the shift of the equilibrium was determined from one bond or long range carbondeuterium coupling constants. In 1 -deuteriosemibullvalene [45] the deuterium shows the same preference for bonding to the aliphatic C-5 bridgehead position over the cyclopropyl position as was found in 1-d,-barbaralone [38]. The equilibrium constant is K = 1.094 - 1.144 between +29 and -44°C giving AW = -84.3 cal mol-' and A P = -O.IOcalmol-'K-'.

SH3

D \

In I ,5-dimethyl-2-deuterio-semibullvalene [46] the isotopic equilibrium (43) was found to be shifted in favour of the isotopomer with deuterium bonded to a cyclopropyl carbon and thus confirms the direction of the equilibrium shift in 2-D-barbaralone [40]. The isotope effect data are K = 1.109-1.191 between +29 and -53"C, AW = -115.9calmol-' and AS" = -0.17calmol-' K - ' . An equilibrium isotope effect was also observed, when the deuterium is not directly attached to the averaged carbons but is two bonds removed. I-Trideuteriomethyl-5-methylsemibullvalene [47] (R = H) ( K = 1.01 I1.021 between +29"C and -86"C, AW = -9.6calmol-' per D and


97

AS' = -0.01 cal mol-' K - ' per D) and the corresponding 2,4,6,8-tetramethylcarboxylate [47] (R = COOCH,) ( K = 1.015-1.035 between +29"C and -92°C AH" = -5.9cal mol-' per D and A 9 = -0.01 cal mol-' K - ' per D) show the same direction of perturbation of the Cope equilibrium (44) as in I-deuterio-semibullvalene [45] although the effect is much smaller. The data for the tetramethylcarboxylate [47] (R = COOCH,) may be less accurate because the static shift difference for the frozen out species may be different from that in [44]. The mono- and di-deuteriated 1,5-dimethylsernibullvalenes [47] have correspondingly smaller isotope shifts and thermodynamic parameters indicating that the isotope effect is approximately additive. Cyclohuturiiene A direct spectroscopic proof of the valence isomeric rectangular structure of cyclobutadiene [48] was obtained from the investigation of 1,2,3-tri-t-butylcyclobutadiene [49] in which one outer t-butyl group was perdeuteriated (Maier et ul., 1982). The peak for the doubly populated olefinic site (C-I, C-3) in [49] shows a temperature dependent splitting of 0.4534.297 ppm between -96°C and -62"C, indicating isotopic perturbation of a fast equilibrium (46) of valence isomers. Carbon C-l substituted with the deuteriated t-butyl group is shifted upfield.

R RQR

H

R]m(

R

H

The model compounds 2,3,5-tri-t-butylcyclopentadienone [50] and 2,3,4-tri-t-butylcyclopentadienone[ 5 I ] show that in [50] an upfield shift of 8.5ppm is observed for C-3, which is connected to the other t-butyl substituted olefinic carbon (C-2) via the double bond, compared to [51], where C-4 is connected to the t-butyl substituted olefinic carbon C-3 via a single bond.

HANS-ULLRICH SIEHL

98

0

Rx; 0

R [jOI

[ j11

The upfield shift of C- 1 in [49] shows that the isomer [49b] in which C- 1 is connected to C-2 via the double bond is favoured in the equilibrium. The equilibrium constant K = 1.1 13-1.073 between -96°C and -62°C; values of AH" = -84.6calmol-' and A F ' = -0.26calmol-' K - ' were calculated using 8.5 ppm as shift difference for the frozen-out static structures. In accord with the vibrational origin of equilibrium isotope effects, the observed direction of the equilibrium shift can be rationalised as follows. Steric hindrance of two adjacent t-butyl groups connected via the shorter bond, i.e. the double bond, is more pronounced than steric hindrance of those t-butyl groups which are connected by the longer single bond. The potential energy for C-H/C-D vibrations in the t-butyl groups is perturbed by this hindrance. Vibrational frequencies and the zero point energy for C-H/C-D vibrations in the t-butyl-groups connected via the shorter bond are higher. The deuteriated t-butyl group is raised in energy in the sterically more constrained bonding situation less than a protiated t-butyl group; hence the equilibrium is shifted towards [49b]. CONFORMATIONAL EQUILIBRIA

1,I ,3,3-Tetramethylcyclohexane At temperatures where the ring inversion (47) is fast, 1,1,3,3-tetramethylcyclohexane shows only one averaged methyl signal in its 13Cnmr spectrum. The degeneracy of the fast conformational equilibrium is lifted in [52] which has one deuteriated methyl group. This leads to a splitting of the averaged

cH3Q

(47)

resonances and in the methyl region three lines are observed in the high field 13C nmr spectrum of [52] (Saunders et al., 1980a). Two peaks separated by 0.184 ppm were assigned to the two methyl resonances at C-3. Since axial methyl groups absorb at higher field than equatorial methyl groups, the high

ISOTOPE EFFECTS

ON

NMR SPECTRA

99

field line was assigned to the C-3 methyl group which spends more time in the axial position and conversely the low field line represents the C-3 methyl group which spends more time in the equatorial position. A broader peak in between the two outer resonances arises from the CH,-group at C-I. The line-broadening is due to unresolved long range deuterium couplings. This peak is shifted 0.08 ppm upfield from the lowfield C-3 methyl peak as a result of the intrinsic upfield shift of the geminal CD,-group. The fact that this line is upfield of only that one of the two C-3 methyl peaks which was assigned to the averaged equatorial methyl groups shows that CH, at C-I is preferred in the equatorial position and the geminal CD,-group in the axial position. The equilibrium (47) is shifted in favour of conformation [52b]. The CD,-group at C-l should show a large one bond upfield intrinsic isotope shift and should appear about 0.5ppm upfield of the CH,-signals but it was not observed under the experimental conditions. The equilibrium constant K = 1.024 at 17°C was calculated using 6 = 0.184 ppm for the equilibrium isotope splitting and A = 9.03 ppm for the shift difference of the axial and equatorial C-3 methyl groups at slow exchange ( - IOOOC). The free energy difference is AGO = - 24 cal mol- in favour of the conformer with the axial CD,-group. The reason for the conformational preference of the CD,-group in the axial position is that 1,3-interactions between two axial methyl groups are sterically more hindered than in the 1,3-diequatoriaI arrangement. The hydrogen vibrations in the axial methyl groups are thus more closely confined resulting in a steeper and more curved potential energy well compared to the equatorial situation. This leads to greater force constants, increased C-H vibrational frequencies and higher zero point energy for the axial methyl groups. The CD,-group has a lower zero point energy compared to CH,-groups and thus is raised less in energy in positions which have higher zero point energy. The axial preference for the CD,-group gives the lowest total zero point energy and lowest total energy. Substituted piperidines, cyclohexanes and dioxanes Similar deuterium isotope effects on the conformational equilibrium (48) in N-trideuteriomethyl-N,3,3-trimethylpiperidiniumion [53] were studied by

CH3

CH3

~ 3 1

100

HANS-ULLRICH SlEHL

Robinson and Baldry, (1977a). The deuteriated methyl group is preferred in the axial position. The equilibrium constant at 27°C is K = 1.028 giving a free energy difference AC" = - 16.7 cal mol-'. It was also shown that the syn 1,3- and IS-diaxial interaction of one methyl group at C-l with two hydrogens at C-3 and C-5 is sufficient to observe a conformational equilibrium isotope effect in N-trideuteriomethyl-N-methylpiperidinium ion [54] and in the deuteriated trans1,3-dimethylcyclohexane [55] and rrans-2,6- and trans-3,5-dimethylcyclohexanones [56] and [57] (Robinson and Baldry, 1977b).

In these cases the equilibrium constants are not far from unity ( K = 1.01-1.03 and AGO = - 19 to - 11 calmol-'. Uncertainties in the 13C-spectralmeasurements, which were done at 23 MHz, and possible errors

ISOTOPE EFFECTS

ON NMR SPECTRA

101

in the estimation of static shift differences and intrinsic isotope shifts were discussed; they preclude in part comparison of these small effects. In cis-I-ethyl-4-trideuteriomethylcyclohexane [58], where the conformational equilibrium (53) is non-degenerate ( K H= 1.08 at 25°C) in the unlabelled

-

p - " D L c ~ : H ~

y 2

(53)

H

CD3

CH3 1581

compound, the same order of magnitude for the deuterium isotope effect was observed using high field 13C nmr spectroscopy (Booth and Everett, 1980b). The ethyl-CH, was used to trace the equilibrium isotope effect because this carbon is six bonds removed from deuterium and shows no intrinsic isotope shifts and also has a reasonably large shift difference at slow exchange which makes this position particularly sensitive to any change in the position of the equilibrium. The isotope effect, K = KD/KH = 1.02 and AGO = - 1 I .9 cal mol- ' at 25°C indicates a slightly greater preference for the conformer with axial CD,-orientation compared to axial CH, orientation in the unlabelled compound. In [l3C-l-methyl]-cis-1,4-dimethylcyclohexane [59] no isotopic perturbation was observed within experimental error. It was concluded that a 3C-isotope effect on the degenerate equilibrium (54) in cis-l ,Cdimethylcyclohexane is either too small to be detected or non-existent (Booth and Everett, 1980a).

'

1591

The conformational equilibrium of trans-3-methyl- and 3-hydroxymethyl- I-trifluoromethylcyclohexane is shifted at room temperature in favour of the conformer with an equatorial trifluoromethyl group. Isotopic perturbation of these nondegenerate equilibria ( 5 5 ) was observed in the deuteriated compounds [60] (R = D or OH) using CF, as a reporter group to monitor the isotope effect by "F-nmr spectroscopy (Robinson and Baldry, 1977b). In [60]the "F-resonance is shifted further upfield compared to the corresponding unlabelled compounds showing that the equilibria are shifted further in favour of conformations [60b] with the CF,-group in the

HANS-ULLRICH SIEHL

102

equatorial position. The deuteriated methyl- and hydroxymethylene-groups are found more frequently than the nondeuteriated groups in the axial position which has a higher zero point energy than the equatorial position due to more confined C-H(D) vibrations caused by the syn-axial interactions with the hydrogens at C-l and C-5. The quantitative interpretation of the small effects is somewhat limited by experimental errors involved in the measurements at low field and the uncertainty of the estimation of static shift differences. The calculated isotope effects per deuterium were found to be constant for successive deuteriation within experimental error. This is consistent with the interpretation of conformational equilibria in cyclohexanes that the strain associated with an axial group -CHXY is largely attributable to repulsion between the hydrogen in -CHXY and the syn IJ-diaxial ring H-atoms. 2,6,6-Trideuterio-2-methylcyclohexanoneshows isotope effects on the 3C chemical shifts which could not be explained assuming intrinsic effects only (Wehrli and Wirthlin, 1976). Saunders et a1 (1980a) have suggested that the origin of these effects is an equilibrium isotope effect. Conformational isotope effects have been observed in deuteriated 1 ,Cdioxanes (Jensen and Neese, 1971) and in substituted 1,3-dioxanes (Robinson, 1971). A long range intrinsic deuterium isotope shift in 2,2-bis(trideuteriomethyl-5,5-dimethyl-1,3-dioxane [6 I ] was observed by Anet and Dekmezian (1979). This molecule does not show an equilibrium isotope effect because the two rapidly interconverting chair isomers have exactly the same energy. 2-D-5.5-Dimethyl- I ,3-dioxane, however, exhibits an isotope effect on the conformational equilibrium (Anet and Kopelevich, 1986a). Deuterium is preferred in the equatorial position (AGO = -49cal mol-’ at 25°C).

CH3 [611

cis-2-Decalone The conformational equilibrium (56) between the “steroid-like’’ [62a] and “non steroid-like” [62b] conformation of cis-2-decalone is shifted in favour


103

of [62b] because this conformation has one less hydrogen-mediated gauche butane type interaction (3-alkylketone effect).

[63a]

~3b1

The 3C nmr spectrum of P,P’-tetradeuteriated cis-Zdecalone [63] shows intrinsic isotope shifts over up to four bonds as well as equilibrium isotope effects over five bonds (Lippmaa et al., 1982). Under fast exchange conditions at room temperature, no net isotope shifts are observed at carbons C-7 and C-9, which are separated from the deuterium by four bonds, because the intrinsic isotope shift and the equilibrium isotope effect on the chemical shift happen to have the same magnitude but different sign, thus compensating each other. At -70°C the equilibrium is frozen out and consequently equilibrium isotope effects disappear. The remaining four bond intrinsic isotope shifts for C-7 and C-9 are 0.02 ppm downfield and are only observed in conformation [63a] indicating the presence of deuteriummediated gauche interactions to C-4 and C-2 which are not present in [63b]. The equilibrium isotope shifts in the fast exchange spectra must be 0.02 ppm upfield for C-7 and C-9 to give the apparent zero effect at room temperature. This direction of the equilibrium isotope shift is towards the chemical shift of C-7 and C-9 in [62a] which are upfield compared to [62b]. The equilibrium isotope effect is observed unmasked by an intrinsic isotope shift at C-8 which is separated by five bonds from deuterium. At room temperature, C-8 is shifted 0.02 ppm to lower field and has an identical shift to the unlabelled compound in the frozen-out spectrum at - 70°C. This confirms the equilibrational origin of the isotope shift of C-8 observed under conditions of fast exchange. At -70°C the C-8 carbon appears at 26.6 ppm and at 20.5 ppm in conformation [62a] and [62b] respectively. The fact that the averaged chemical shift of C-8 is more downfield in the deuteriated compound [63]

HANS-ULLRICH SIEHL

104

compared to the unlabelled ketone [62] shows that the equilibrium in [63] is shifted towards conformation [63a] where the C-D vibrations are more confined due to one additional gauche butane type interaction as compared to conformation [63b]. C~*clodec.anotie Deuterium isotope shifts over up to six bonds have been observed in the 13C nmr spectrum of deuteriated cyclodecanones (Wehrli et a/., 1978). The equilibrational origin of the observed long range effects was briefly discussed by Anet and Dekmezian (1979) and was explained in detail by Whipple et a/. (1981). In the low temperature spectra of deuteriated cyclododecane isotopomers Anet and Rawdah (1978) have also observed deuterium isotope effects which are likely to have a conformational origin and to arise from the lack of precise D,-symmetry in the preferred conformation of that hydrocarbon. Low temperature nmr measurements have shown that cyclodecanone [64] has an unsymmetrical boat-chair-boat (BCB) structure which has four transannular H-H interactions (Anet et al., 1973). At room temperature all 13Cchemical shifts except the carbonyl carbon and C-6 are a 1 : 1 average of the shifts of the two equivalent conformations [64a] and [64b]. In cyclodecanone deuteriated at C-6, the symmetry of the conformational equilibration process (58) is not perturbed and hence no shifts arise as a result of unequal weighting; but intrinsic isotope shifts over up to three bonds are observed.

[64a]

[64bI

In contrast, deuteriation at C-2 lifts the degeneracy of the conformational equilibrium, and, in addition to intrinsic isotope shifts over up to three bonds, equilibrium isotope splittings are observed for the carbons pairs C-2/C-10, C-3/C-9, C-4/C-8 and C-5/C-7 which are averaged in the conformational equilibrium. If corrections for intrinsic isotope shifts are taken into account, the splittings are symmetrical with respect to the unperturbed position in the nonlabelled compound within experimental error. The sign of the equilibrium isotope effect can be extracted from the corrected relative upfield and downfield shifts of the perturbed peak positions in 2-D,-[64] compared to the corresponding equally averaged position in [64]. Upfield shifts for positions 9 and 4 and downfield shifts of equal magnitude for positions 3 and 8 were observed in 2-D,-[64] with respect to the corresponding C-9/C-3 and C-4/C-8 resonances in [64]. The C-9 and C-3 and also the C-4

ISOTOPE EFFECTS

ON NMR SPECTRA

105

and C-8 peaks in 2-D,-[64] could be assigned individually on the basis of deuterium coupling to C-3 and C-4. Whipple suggested a reversal of the assignment for C-5 and C-7 in [64] which would also give an upfield shift for C-7 in 2-D1-[64]. The relative upfield shift for carbons C-4, C-7 and C-9 is in accord with a preference for 2-D, -[64b] in equilibrium (58). Transannular interaction between protons at positions 2, 6 and 9 and between the protons at positions 4 and 7 result in an upfield steric compression shift for this carbon. The C-D vibrations for deuterium at C-2 are more confined and thus have higher zero point energy in conformation [64b] than in conformation [64a] where transannular interactions occur between positions 3, 6 and 10 and between positions 5 and 8 and do not involve the deuterium at C-2. In accord with this interpretation smaller effects were observed in cyclodecanone deuteriated at C-5 which has only one transannular interaction. Similar conformational equilibrium isotope effects are likely to be the cause of the upfield and downfield long range isotope shifts observed in monodeuteriated cyclododecanones (Jeremik et a/., 1982). Cyclohexane Giinther and Aydin (198 I ) have investigated the conformational equilibrium (59) in d,-cyclohexane. The two conformational isomers [65a] and [65b] with deuterium in the axial and in the equatorial position can be frozen out in the 100 MHz I3C spectrum below -80°C. The deuteriated carbons in the two isomers have different intrinsic isotope shifts which are upfield compared to the nondeuteriated carbons. The triplet with the smaller 'JCDcoupling constant was assigned to the carbon with axial deuterium and was shifted 0.0482ppm to higher field than the triplet caused by the carbon with equatorial deuterium.

[65a]

At room temperature the averaged signal for the deuteriated carbons was observed slightly (0.01148 ppm) downfield from the average value expected for an unperturbed equilibrium. This indicates that K # 1 and that the equilibrium (59) is shifted in favour of conformation [65b] with deuterium in the equatoriai position. The small effect was confirmed by two independent experimental observations. In a mixture of Do-, D,- and I , I-D,-cyclohexane, the deuteriated carbon in D,-cyclohexane showed an upfield shift which was 0.01689ppm, less than half the value of the intrinsic upfield shift in

106

HANS-ULLRICH SIEHL

1 ,I-D,-cyclohexane. The nonadditivity of the isotope shifts shows that an equilibrium isotope effect is superimposed on the intrinsic isotope shift in D,-cyclohexane. Integration and line shape analysis of the two signals for the axial and equatorial deuteriums in the 61.4 MHz 'H nmr spectrum of [65] at -88°C showed the expected population change and led to the same conclusion. At room temperature the equilibrium constant K = 1.100 and AGO = -48 cal mol-' in favour of conformational isomer [65b] with deuterium in the equatorial position. Vicinal gauche H,H(D) interactions were suggested to be the cause of the observed conformational preference. Equatorial hydrogens in cyclohexane are gauche to both adjacent axial and equatorial hydrogens and the axial hydrogens are gauche to both vicinal equatorial hydrogens but anti to two vicinal axial hydrogens. The C-H and C-D vibrations are thus more confined in the equatorial positions. Therefore the equatorial C-H(D) bonds have higher vibrational force constants which in turn result in higher zero point energy for these positions. A priori, C-D vibrations have lower zero point energy compared to C-H bonds and the energy decrease is greatest when hydrogen is favoured in the axial and deuterium in the equatorial position. Recent calculations of the conformational equilibrium isotope effect for [65] using a scaled quantum mechanical 3-21 G force field are in qualitative accord with the experiment and confirm the zero point energy origin of the effect (Williams, 1986). The smaller force constants for axial C-H than for equatorial C-H bonds are suggested not to be caused by less vicinal H,H steric interactions for axial hydrogens. Instead they may reflect a greater o-o* interaction between an antibonding orbital of an axial C-H bond and the bonding orbitals of a pair of antiperiplanar axial C-H bonds than the interaction between an antibonding orbital of the equatorial C-H bonds and a pair of antiperiplanar C-C bonding orbitals. The same direction for the isotope effect on the ring inversion equilibrium in trans-( 1,4-protio)-D,,-cyclohexane has been observed using 200 MHz 'H nmr spectroscopy (Anet and Kopelevich, 198613).The (aa)-conformation with the protons at C-l and C-4 in the axial position is favoured in the equilibrium. At 25°C an upfield shift of 0.00255 ppm was observed relative to the corresponding cis-( 1 ,4-protio)-isomer which exhibits no conformational isotope effect (Kcis= l ) due to symmetry of the ea- and ae-isomers. Using the equation K = (A 26)/(A - 26) with A = 0.478 ppm the equilibrium constant K = 1.022 is obtained which yields AGO = -6.3 cal mol-' for a single deuterium. These values are significantly smaller than those for [65]. Molecular mechanics model calculations on D, ,-cyclohexene using stretching force constants for equatorial and axial C-H bonds which were

+


107

adjusted to reproduce experimental vibrational frequency differences are in agreement with the experimental nmr results. If vicinal H,H(D) interactions are determining the conformational equilibrium isotope effect, the results can be rationalised in the following way: C-D bonds have lower vibrational amplitude than C-H bonds and therefore it might be anticipated that vicinal gauche interactions of CD,and CHD-groups in D,,-cyclohexane are smaller, resulting in smaller perturbations than the vicinal gauche interactions between CH,- and CHD-groups in D,-cyclohexane [65]. I t is difficult, however, to compare small isotope effects obtained by different experimental methods. The data for [65] were obtained over a temperature range of more than 100°C using the peak area method at slow exchange and the chemical shift difference for the averaged intrinsic isotope shift under conditions of fast exchange to determine the equilibrium isotope effect. For trans-( 1,4-protio)-D,,-cyclohexane the chemical shift difference method was applied but no temperature dependent data were reported; thus an important diagnostic criterion to characterise an equilibrium isotope effect is not accessible. In heavily deuteriated compounds additional problems arise in defining precisely the contribution of intrinsic isotope shifts to the observed isotope splittings. Cycloheptatriene and cyclopentene The degeneracy of the rapid conformational equilibrium between the nonplanar conformations of cycloheptatriene is lifted in 7-deuteriocycloheptatriene [66]. The equilibrium constant is no longer unity but conformer [66b] with hydrogen syn to the ring is present in higher concentration (Jensen and Smith, 1964).

[66a]

Two separate peaks were observed for the CH,-protons in cycloheptatriene below -141°C. From a consideration of the different coupling constants caused by different dihedral angles between the two different methylene protons and the vicinal vinyl hydrogens, two peaks at 1.44 and 2.88 ppm (at - 170°C) were assigned respectively to the methylene hydrogens in positions syn and anti to the ring. The area of the peak of the syn hydrogen was larger than that for the anti hydrogen and the ratio was found

108

HANS-ULLRICH SIEHL

to increase in favour of the peak for the syn-hydrogen with decreasing temperatures, indicating that the conformer [66b] with syn-H and anti-D is lower in energy. At fast exchange the shift of the peak for the averaged methyiene protons was found to be temperature-dependent, whereas the shift of the vinyl hydrogens was independent of temperature. When the temperature was lowered, the methylene resonance moved upfield which is towards the position of the .yvn-H resonance. The area ratio at slow exchange and the direction of the equilibrium isotope shift at fast exchange led to the same conclusion that conformer [66b] is favoured in equilibrium (60). The steric effect of the eclipsing interaction between the anti-position at the methylene carbon and the two adjacent vinylic hydrogens confines C-H(D) vibrations in this position more than in the syn-position. Deuterium is therefore preferred in the anfi-position and hydrogen in the synposition. The reported values for the isotope effect are K = 1.41-1.10 with AGO = -72 to -3Ocalmol-'between -168°C and -114°C; AH" = -142 k 30calmol-' and AS" = -0.7 k 0.3calmol-'K-'. The isotope shifts observed in the 'H nmr spectrum of cis-3,5-dideuteriocyclopentene [67] (Anet and Leyendecker, 1973) have their origin most probably in conformational equilibrium isotope effects and could be explained in a similar way to the effects in cycloheptatriene, although the size of the effects is different because of different conformational arrangements.

H

BRIDGING A N D HYPERCOORDINATION IN TRANSITION-METAL COMPLEXES

Coordinatively unsaturated species can achieve saturation through partial bonding to the hydrogen or carbon atoms of organic ligands. Metalhydride-metal and metal-hydride-carbon interactions in transition-metal complexes play an important role in catalytic reactions like carbon monoxide reduction, olefin metathesis, alkyne polymerisation and methylene transfer.

Osmium cluster complexes The nature of the hydrogen bonding in methylated p-hydridodecacarbonyltriosmium cluster compounds [68] and [69] has been investigated by


109

Shapley and coauthors using neutron diffraction (Shapley et al., 1979) and nmr spectroscopy in combination with isotope effects (Shapley and Calvert, 1977, 1978; Shapley et al., 1978).

Reaction of di-p-deuteridodecacarbonyltriosmium with diazomethane provided a mixture of hydridomethyl- and hydridomethylene-tautomers of Os,(CO),CH,D,, D,-[68] and D,-[69], which interconvert slowly in solution. The ' H nmr spectra show a preference for H over D in the metal hydride sites. An isotope effect on the slow exchange equilibrium (61) favours incorporation of deuterium in the methylene- and methyl-positions and hydrogen in the osmium hydride positions. The effect has been interpreted in terms of vibrational zero point energy differences. The C-H bonds have significantly higher vibrational frequencies than the metal hydride bond. Partial substitution by deuterium leads to preferential placement of the lighter isotope in the lower frequency site. Estimation of the corresponding frequencies from vibrational spectra led to a range of calculated equilibrium constants for the H/D incorporation which are consistent with the experimentally observed results. The large disparity in zero point energies found in this case was considered to be generally characteristic of a fully equilibrated H/D distribution among carbon (or nitrogen or oxygen) and metal sites. Isotopic perturbation of the 'H nmr spectra was found for the shielded ( - 3.68) methyl signal in partially deuteriated decacarbonylhydridomethyltriosmium (Shapley and Calvert, 1978). The spectrum shows separate CH,D- and CHD,-signals displaced significantly to higher field from the CH,-signal of the nondeuteriated compound [68]. The separations vary strongly with temperature, increasing from 0.34 ppm (CH,D) and 0.39 ppm (CHD,) at 35°C to 0.55 and 0.68 ppm at -76°C. These large temperaturedependent shifts are inconsistent with the relatively small intrinsic isotope shifts commonly observed upon geminal substitution of H by D (0.01 ppm) and indicate an isotopic perturbation of a fast equilibrium. No line broadening was observed even at - 100°C. This sets an upper limit of approximately 5 kcal mol- for the barrier of the exchange process.

HANS-ULLRICH SIEHL

110

H I \ [70a]

A model (62) involving C---H-0s interactions can rationalise the observed effect. For the monodeuteriated methyl-compound [70] three structures [70a-c] are possible.2 An isotope effect on a fast equilibrium between these structures is to be expected. The lower vibrational frequency and hence lower zero point energy of the bridging C-H bond leads to a preference for the lighter nucleus at the bridging site, and both H-bridged forms [70b] and [~OC] will be slightly more abundant than the D-bridged form [70a]. Since the bridging atom should resonate at higher field than the terminal hydrogen atom, the nonrandom distribution results in a net upfield shift for the CH,D-signal relative to the CH,-signal of the unlabelled compound. As expected for an equilibrium isotope effect, the isotope shifts increase when the temperature is lowered because the equilibrium is shifted in favour of the lower energy (H-bridged) sites, [70b] and [~OC]. Neglecting a diastereotopic shift difference of the two terminal hydrogens which is only about 0.2 ppm in analogous static compounds, the exchange process is threefold degenerate, taking place between a doubly populated lowfield site and a singly populated highfield site. Analysis of the temperature dependence of the averaged shifts in the CH,D- and CD,H-complexes led to bridging hydrogen and calculated shifts of - 15 f 1 ppm for the C-H-0s 2 f 1 ppm for the two terminal hydrogens. The energy difference between the D-bridged [70a] and H-bridged [70b] and [~OC]was found to be AH" = - 130 cal mol- Within experimental error identical results were obtained for the mono- and di-deuteriated species. Support for this interpretation comes from the temperature-dependent isotope effect on the size of the averaged lJCHcoupling constant. Whereas

'.

For clarity, simplified formulae are used in this section for structures [70a

(94)

[ I34bI

The o-delocalised bicyclobutonium ion structure was suggested to be a set of unsymmetrical cations [I321 by Olah et a(. (1970). The possibility of a symmetrical single minimum structure [ 1331 (i.e. [ 1301 or [ 1391 respectively) was also considered but claimed to be indistinguishable from a double potential energy minimum with a very low barrier (Olah et al., 1978a). The unsymmetrical structures [I321 which are still depicted as in (94) (Olah et al., 1985c), imply three different CH,-groups (Olah et al., 1972); in fact three different chemical shifts for the three CH,-groups were used to calculate the anticipated averaged position in the bicyclobutonium ion [ 1391 (Olah et al., 1978b). This would require a three-site, singly populated, fast exchange at intermediate temperatures which is not consistent with the observed pattern for the isotope splittings in the C4H,DCH3+ cation. Equilibrium isotope effects depend only on force constant differences of C-H bonds, which are to be expected for the CH,-groups in the envisaged exchanging structures [ 134al and [134b] and are independent of the energy barrier between these two structures. The absence of equilibrium isotope effects in the very low temperature spectra of the methylene-deuteriated bicyclobutonium ions rules out any equilibrium between the unsymmetrical structures [134a] and [134b] and provides convincing evidence that the symmetrical methylbicyclobutonium ion [ 1301 is the minimum energy structure for cation [123]. Similar conclusions have been drawn recently by Olah

HANS-ULLRICH SIEHL

142

et al. (1985b). who have also observed isotope effects in CD,-labelled C,H,CH, . +

Bicylobutonium cation Evidence for equilibrating cations and stereospecijic isotope effects Even at the lowest temperatures the 13C nmr spectrum of the C,H,+ cation, unlike that of its substituted homologue C4HSCH3+,has only a single peak for the three methylene carbons, indicating real or time-averaged three-fold symmetry. The CD,-labelled cation C4H,D2 shows a temperature-dependent upfield shift for the nondeuteriated methylene carbons of 1.77 ppm ( - 135°C) to 1.24 ppm ( - 107°C) (compared with the protio-ion), indicating a definite equilibrium isotope effect. A three-fold symmetrical static structure like the tricyclobutonium [135] is not consistent with this result and cannot be the main species present (Saunders and Siehl, 1980). +

H I

H [ 1351

The C,H,+-cation has two different sets of three averaged methylene protons. Mono- and trideuteriated cations prepared from 1 -D,-cyclopropylmethanol [ I361 (Saunders and Siehl, 1980) or from l-D,-(2’,2’-D2-cyclopropy1)methanol [I371 (Roberts et a/., 1984) are a mixture of stereoisomeric cations with the CHD-deuterium in the low field and in the high field methylene positions. When (2’E, 2’2, 3’E)-D,-cyclopropylmethanol [ 1381 is used as a precursor (Roberts et al., 1984), only one stereoisomeric C,H,D,+ is observed. The deuterium at the monodeuteriated methylene carbon is in the endo position, i.e. truns to the methine proton. From the intensity ratio of the signals for the remaining two em-protons and the single endo-proton, the lowfield signals could be assigned to the endo- and the highfield signals to the em-protons. The interconversion of exo- and endo-stereoisomers [ 1391 is slow on the nmr timescale at - 90°C. This indicates that inversion via [ 13I], which is a tertiary cation for [130], is unfavourable for [I391 because a secondary cyclobutyl cation would be involved. The alternative pathway via rotation about a methylene to methine carbon bond also has a high activation energy.

143


D

CHDOH

-

+

C4H6D

exo and endo

+

C4H4D3 exo and endo

[ I381

The proton spectra of a mixture of stereoisomeric D,- or D,-cations show different equilibrium isotope effects for the endo- and exo-deuteriated cations. When the deuterium at the CHD-methylene group is in the endo(lowfield) position, large isotope splittings between 0.388 ppm ( - 145°C) and 0.142 ppm ( - 55OC) in C,H,D+ and 0.29 ppm ( - 121°C) in C,H,D,+ were observed for the exo-(upfield) protons but there was no measurable effect on the remaining endo-protons. The stereoisomeric C,H,D, +-cation with the CHD-deuterium in the exo-position showed no significant isotope effect and, e.w-D-C,H,D+, only a small temperature-dependent downfield shift was observed for the em-protons and there was no effect on the endo-protons. In the proton spectra of the geminally dideuteriated cation C,H,D,+, the upfield methylene protons were found to be shifted to higher field (relative to the protio-ion) by between 0.087 ( - 130°C) and 0.057 ppm (- 80°C) and the downfield methylene peak was unaffected. This observation, that the equilibrium isotope shifts for endo- or exodeuterium are sizeable for the high field averaged exo-protons only, but negligible for the low-field averaged endo-protons can only be rationalised by assuming that equilibrium among nonequivalent methylenes is perturbed by deuterium, but that the chemical shift difference between the rapidly equilibrating hydrogens, if the fast interconversion process could be stopped, is much larger for the exo-protons than for the endo-protons. The isotope effects in the two C,H,D+-cations give additional information on the dynamics and structure of the C,H,+-cation (Saunders and Siehl, 1980). The intensity pattern observed in the proton spectrum shows the multiplicity of the exchange and the relative populations of the exchanging sites. The direction of the isotope shifts can be assigned to the exo- and

144

HANS-ULLRICH SIEHL

endo-deuteriated cations using the exolendo-proton assignment of Roberts et al. (1984). The sign of the equilibrium isotope effect and the relative C-H bond force constants at the exchanging sites can be determined. In the endo-D,-cation, a signal of intensity 1 for the CHD-exo-proton moves about twice as much to lower field as a signal of intensity 2 for the two CH,-exo hydrogens moves upfield compared to the protio-ion and corrected for geminal intrinsic isotope shift. The relative size of the shifts indicates that a two-site exchange between a doubly populated low field site and a singly populated high field site is perturbed by deuterium. The intensity ratio of the two peaks for the exo protons is 1 : 2, one exo-CHD to lowfield and two exo-CH, to high field. This shows that CH,-endo-protons are preferred at the upfield site. The upfield endo-protons have lower vibrational bond force constants than the low field endo-protons. In the stereoisomeric D,-exo-cation, the two remaining exo-protons give a peak of intensity 2 which is close to the unlabelled position but shifted downfield. This indicates that the force constants for the exo-protons at the downfield site are lower than those at the upfield site. The 13C nmr spectrum of the mixture of exo- and endo-C4H,DC cations convincingly confirms the observation of two different isotope effects. As in the proton spectrum, the methylene resonance in the endo-D-cation showed the larger isotopic perturbation. With respect to the protio-ion, the CHD-carbon moved downfield and two CH,-carbons moved upfield. The shift between them varies with temperature from 7.05 ppm ( - 133°C) to 3.82 ppm ( - 87°C). The exo-D-cation showed smaller isotope splittings between 3.16 ppm ( - 1 18°C) and 2.78 ppm ( - 87°C) in the opposite direction, with the CHD-carbon moved upfield and the CH,-carbons moved downfield. Analogous 3C-shift patterns, although with different intrinsic isotope shifts superimposed, were observed by Roberts et al. (1984) in the C4H4D3+-cations. The isotopic perturbation for the CHD- and CH,-resonances at -95°C are 5.34ppm for the CHD-endo-cation and 3.17ppm in opposite direction for the CHD-exo-cation. Additional support for the opposite sign of the equilibrium isotope effects in exo- and endo-d-C,H,D+ comes from the observation that the upfield shift of the CH,-carbons in CD,-labelled C4 H5 D 2 +(compared to the protio-ion) is the algebraic sum of the upfield and downfield shifts of the CH,-carbons in the two monodeuteriated cations. C-H bond force constants and structure Despite the fact that the fast averaging process which renders the methylene carbons equivalent cannot be frozen out, the isotope shifts give information on the relative stiffness of C-H bond force constants in a frozen-out


145

structure. The endo-protons at the single methylene carbon which would give an upfield peak in the spectrum of a static structure have much weaker vibrational force constants than the endo-protons at the two methylene carbons which would give a downfield peak. The exo-protons at the high field methylene carbon have moderately stiffer bond force constants than the exo-protons at the two methylene carbons which would cause the downfield peak in the spectrum of the frozen-out cation. The much larger shift difference for the averaged exo-methylene protons than for the endo-protons suggests that the two protons at the upfield methylene carbon have very different chemical shifts and that the exo-proton at this carbon is located in a unique environment compared to the other protons. H

[I”

These two observations exclude cyclopropylmethyl and cyclobutyl as the dominant structures in the equilibrium of C,H,+-cations. In a bicyclobutonium ion [139], the pentacoordinated carbon is likely to be upfield as for the methyl-homologue. The vibrations for the endo-C-H bond at the pentacoordinated carbon are less confined compared to the other endo-CH bonds because bridging could drain electron density out of this bond. M I N D 0 / 3 calculations of Dewar and Reynolds (1984) have shown a larger bond length and some involvement in multicentre bonding only for this endo-C-H bond. The reverse but smaller isotope effect observed for the exo-protons at the pentacoordinated carbon must be caused by more hindrance to vibration compared to the other e.w-hydrogens. This may be due to greater crowding at the pentacoordinated carbon and 1,3-hydrogen interaction between the 3-exo- and the I-methine-protons. The unique shift of the eso-proton was suggested by Roberts to be caused by its apicai position at the pentacoordinated carbon which could possess a configuration with the elements of square-pyramidal geometry. Both directions for the equilibrium isotope effects have their counterparts in kinetic isotope effects. The situation for the Pndo-proton resembles the transition state in an S,2 reaction with an a-primary kinetic isotope effect (kH/kD> I). The endo-proton would be the leaving group with a loose bond, and the bridging cationic carbon an incoming electrophile. The confined

146

HANS-ULLRICH SIEHL

situation for the em-proton can be compared with an a-secondary kinetic isotope effect (k,/k, < 1) which results from increased hindrance to bending vibration in the pentacoordinated transition state. Olah and Roberts (Olah et al., 1978b) interpreted a small temperature dependence of the 13C nmr spectrum of 13C-labelled C,H,+ as indicating a minor species like [I241 in rapid equilibrium with the major ion [139]. A splitting of 0.4 ppm between the methine peaks of two isomeric deuteriated cations observed by Saunders and Siehl (1980) and also by Roberts et ul. (1984) might indicate an isotope effect on a nondegenerate equilibrium in support of this idea. HYPERCONJUGATION IN CARBOCATIONS

C-H H!?percot!jugatioti arid deuteriutn isotope eflects The origin of kinetic P-deuterium isotope effects in solvolysis reactions proceeding through carbocation-like transition states has generally been attributed to C-H hyperconjugation (Sunko and Hehre, 1983). Drainage of electron density from the valence orbitals of C-H bonds into the vacant orbital at the carbocation centre leads to a weakening of all force constants for C-H vibrations and not just those associated with C-H stretching as suggested by a simple hyperconjugative model (Hehre et al., 1983). Vibrational frequencies and zero point energies of p-C-H bonds involved in hyperconjugation are therefore lower than for C-H bonds which are either not involved or are at more remote positions. In an isotopic equilibrium of otherwise degenerate carbocations, where an exchange situation for H and D exists between a hyperconjugating P-position to the positive charge and another position, the equilibrium is shifted to the side where the hydrogens are preferentially bonded at the P-carbon and deuteriums at the more remote carbon position. Saunders and coworkers have applied the isotopic perturbation technique to a number of persistent carbocations and have demonstrated that C-H hyperconjugation is a principal cause of P-secondary deuterium equilibrium isotope effects in stable carbocations. The results fully support the interpretation of secondary kinetic isotope effects in solvolytic substitution reactions. 2.3-Dimethyl- and 2,2,3-trimeth~d-2-hut~d cations. Saunders and coworkers (Saunders et al., 1971; Saunders and Vogel, 1971b) have studied the influence of methyl deuteriation in the 2,3-dimethyl-2-butyl cation using 'H-nmr spectroscopy. This cation undergoes a rapid degenerate hydride shift leading to a single time-averaged doublet for the methyl protons. Substitution of the methyl protons on one side of the molecule by 1 to 6 deuteriums lifts the degeneracy between the two isomers. In the monodeuter-


147

iated cation [ 1401 two new time-averaged methyl doublets are observed. The low field doublet results from averaging of six protons of the methyls attached to the C+-carbon in [140b] with the six protons of the methyl groups attached to the methine-carbon in [140a]. The doublet at higher field has the lower intensity and is due to the averaging of five protons of the methyl groups attached to the methine carbon in [140b] with five protons of the methyls on the C+-carbon in [140a]. The six-proton doublet being located at lower field than the five-proton doublet indicates that the carbocation prefers to be substituted by unlabelled methyl groups, i.e. the equilibrium is shifted towards [ 140bl.

The hydride shift in [I401 is fast even at low temperatures; the shift difference for the two methyl-positions was therefore estimated to be 2.1 ppm using the 2-methyl-2-butyl cation as a model. The isotope splittings 6 for the methyl peaks increase with the number of deuteriums from 0.145 ppm, 0.30 ppm and 0.46 ppm at - 76°C in the D,-D,-labelled cations to 0.623 ppm, 0.779 ppm and 0.928 ppm at - 79°C in the Di-D,-labelled cations. The equilibrium constants were calculated [using K = (A &)/(A - S)] as K = 1.148, 1.33, 1.561 at -76"CfortheD1-D,-cdtions[140]and 1.844.2.18, 2.585 at - 79°C for the D,-D,-[140]. The corresponding enthalpy differences AH" are 54.2, 56.2, 58.3, 58.5, 59.6 and 61.3 cal mol- per D. The symmetrically hexadeuteriated cation [ 11 showed no isotope effect because the two isomers related by the hydride shift (95) are identical and the equilibrium constant is exactly unity.

+

[ 14 I a]

[ I4 I b]

A slow methyl shift equilibrates the unsymmetrical deuteriated ion [I411 with the symmetrical deuteriated cation [I]. This process is affected by the

HANS-ULLRICH SlEHL

148

equilibrium isotope effect on the fast hydride shift (96) in the unsymmetrical deuteriated cations [141a] and [141b]. Cation [141b] with six deuteriums away from the positive charge is favoured over the symmetrical hexadeuteriated cation [I]. An isotope effect of K = 0.954.85 for the slow exchange process from [I411 to [ I ] has been determined from peak intensities of an equilibrated mixture of these cations. The I3C nmr spectrum of cation [I411 shows a very large isotope splitting of 6 = 167.7 ppm at - 138°C for the C+-/methine-carbon pair. The equilibrium constant K = 1.264-1.158 between - 138°C and -71°C and AH" = - 68 f 2 cal mol- per D were calculated using A = 277 ppm as the shift for the C + and C-H carbons (Kates, 1978). These data are probably more accurate than those determined from proton spectra because a larger temperature range was investigated and the assumed A agrees within 1 ppm with the shift difference obtained recently from the solid-state nmr spectrum of the unlabelled cation (Myhre et ul., 1984). The same direction of the equilibrium isotope effect was observed in the nondegenerate 1,2-hydride shift equilibrium of 2-(4'-trifluoromethylphenyl)-3-methyl-2-butyl cation [I421 with one trideuteriomethyl group at C-2 or C-3 respectively (Forsyth and Pan, 1985). The isotope shifts in the I3C spectrum are much smaller (1.3 ppm-1.45 ppm) than in degenerate cations like [141] because K H is very much in favour of the benzylic cation structure for [142].

[ IJ']

CH,,

CD

,'

+ ,CH3 C-CrCH, CH3 [ I433]

CH3, + CH, CH3-C-C,' CD CH3

(97)

f

[IJ3b]

The 2.2.3-trimethyl-2-butyl cation is an example where a rapid methyl shift as in (97) averages all methyl hydrogens leading to one averaged peak in the proton spectrum. When one methyl group is deuteriated in [I431 the equilibrium is shifted towards [ I43bI. The remaining hydrogens continue to give a single averaged peak but this is found shifted downfield from the unlabelled position because the deuteriums disturb the statistical probability of protons residing in the chemically different environment. A value of


149

'

AH" = - 52 cal mol- per D was determined from the temperature dependence of the equilibrium constants ( K = 1.55-1.404 between -92.7"C and - 53°C). Equilibrium constants were calculated from the equation K = (3A + 26)/(3A - 26) using an estimated A of 2.1 ppm as shift difference for the two types of methyl groups (Saunders and Vogel, 1971a). 2,3-Dimeth.~l-2-c~~~clopent~l cation. The equilibrium isotope effects in CD,and P-CD,-dimethylcyclopentyl cations [ 1 161 and [ 1041 have been measured by l3C-nrnr spectroscopy (Saunders et al., 1977a). In [ I 161 the averaged C'/C-H carbons give a pair of peaks split by 81.845.4ppm between - 142°C and -45°C. With K = ( A + S ) / ( A - 6) and A = 261 ppm, K = 1.241-1.124 (per D) and the temperature dependence of K gives AH" = -6Ocalmol-' and AS" = -0.012calmol-'K-' per D. Cation [I041 showed significantly larger isotope effects. The splitting for the C'-/C-H carbons is 105.3-81.6ppm giving K = 1.5341.382 (per D) between - 130°C and -81.6"C and AHo = - I37 cal mol-' and AS" = -0.05calmol-'K-' per D.

(OX)

CH3

CD3

CH3

CD3

[ I lo]

57: F? I L

CH3

CD3

CH3

CD,

I1 441 The direction of the isotope effects on (98) and (99) were determined from proton spectra. In [ I 161 the averaged methyl peak is shifted downfield 0.44-0.34 ppm ( - 93°C to -60' C ) relative to the protio-ion. Cation [I041

150

HANS-ULLRICH SIEHL

showed two averaged methyl peaks with the downfield methyl peak significantly broader due to coupling to methylene protons. The downfield shift for the methyl protons in dimethylcyclopentyl cation [144] which was deuteriated at one methyl group and at the neighbouring methylene group is comparable to the sum of the shifts in [ I 161 and [lo41 indicating a cumulative effect. It also confirms that the isotope effects in [116] and [lo41 have the same relative sign, protons preferring to be at the positions next to the positive charge (Telkowski, 1975). A y-isotope effect in 3,4-dimethyl-3-hexyl cation. The isotope effects observed in cations [ I 161, [I041 and [140]-[144] are differential P- vs y-effects, i.e. Kobserved = K , x K.,. A differential y- vs &equilibrium isotope effect favouring [ 145bl was measured by Kates (1978) in 1 -D3-3,4-dimethyl-3-hexyl cation [I451 ( K = 1.032 at -126°C; AH" = -12calmol-' per D). If one ignores a very small &effect these data can be used to extract the pure P-deuterium equilibrium isotope effects in the cations mentioned above.

Equilibrium C' isotope effects in 2.3-dimethyl-2-hutyl and I ,2-dimethylc.vclopentyyl cations A primary 13C isotope effect on equilibrium (102) was measured by Saunders et al. (1981) in the I3C nmr spectrum of 2,3-dimethyl-2-[2-'3C]butyl cation [6]. The averaged C+/C-H peak was found shifted downfield between 1.35 ppm ( - 135°C) and 0.79 ppm (-62°C) with respect to the corresponding peak in the dilabelled cation [ 1461 which served as a reference since the equilibrium constant for (103) is exactly unity due to symmetry.

CH3,+ c-c v,CH3 CH3'* *\CH 3

CH 7 I

?-c,

CH3'*

+

*

,CH3 CH3

( 1 03)


151

The equilibrium constant for [6], K = 1.0197 (-135°C) to 1.0114 (-62°C) was calculated from the equation K = (A +26)/(A -26) using A = 277ppm. From the van't Hoff equation AH' = -6.0cal mol-' and A T = -0.056cal mol-' K - ' . Any intrinsic isotope shifts were estimated from the precursor alcohol to be smaller than 0.06 ppm. The observed downfield shift shows that the equilibrium favours the positive charge on I3C. The vibrational frequencies associated with the charged carbon must therefore be higher than those of the methine carbon. The more confined bonding situation for the charged carbon despite the smaller number of attached atoms could indicate stiffer C-C bonding caused by C-H-hyperconjugation with the methyl groups.

A secondary 13C equilibrium isotope effect was measured in the 1.2dimethylcyclopentyl cation [I471which was ' 3C-labelled in one methyl group (Saunders et a/., 1977a). In the 13C nmr spectrum of cation [I471 the carbon next to the I3C appeared as a doublet (JCc = 35 Hz) offset downfield between 0.25 ppm ( - 125°C) and 0.10 ppm (-65°C) from the singlet of the other averaged C+/C-H carbon. This shows that the labelled methyl group is preferred next to the charged carbon. Although the C-H frequencies are lower for this methyl group as evident from hydrogen isotope effects in [I 161 and [144], the 13C isotope effect on (104) indicates a small net increase of all vibrational frequencies of the carbon of the CH,-group at the C+-position as compared to the other CH,-group. Both '3C-equilibrium isotope effects in [6] and [147] have the same direction and may be regarded as another piece of evidence for C-H hyperconjugation. In valence bond terminology the bond between the charged carbon and the attached methyl group is stiffer because of partial double bond character. The heavier isotope (' ,C) prefers this more confined bonding, whereas the lighter isotope (I2C) prefers to be at the remote position where the bonding is less stiff. Another secondary '3C-equilibrium isotope effect has been observed in [5] using natural abundance 13C nmr spectroscopy (Olah et a/., 1985a).

HANS-ULLRICH SIEHL

152

The mugni1ude of P-deuterium equilibrium isotope efSects The magnitude of the P-secondary deuterium kinetic isotope effects in solvolyses which proceed viu a cationic transition state have been shown to be conformationally dependent (Sunko and Hehre, 1983). Shiner furnished two important representative examples relating the dihedral angle between an adjacent C-H or C-D bond and the vacant p-orbital of the cationic transition state with the magnitude of the kinetic isotope effect. In the

methylene-deuteriated, conformationally unambiguous, rigid dibenzobicyclo[2.2.2]octane derivative [ 1481, the angle between the C-D bonds and the developing p-orbital of the cationic transition state is approximately 30" and a normal kinetic isotope effect of k,/k, = 1.14 at 45°C was observed. In [I491 the p-C-D bond is essentially orthogonal to the developing p-orbital in the cationic transition state. Shiner and Humphrey (1963) reported an isotope effect of k,/k, = 0.986 k 0.01 obtained from two measurements giving considerably different rate constants, so that the error limit could include k,/k, = 1. This result was often interpreted as a small inverse p-effect but is really a combined p- and y-effect. Further evidence for the conformational dependence of isotope effects comes from the low value of k , / k , = 1.08 for ~-CDZ-2-chloro-2,4,4-trimethylpentane [ 1511 compared to k,/k, = 1.34 for ~-CDZ-2-chloro-2-methylpentane [ 1501 (Shiner, 1961). CH3 /

C H3- C H2- CO 2- C,-C I CH3 [ I51)]

C H3-

CH3 1

/CH3

CH3

CH,

7- CD 2- F-CI llSl]

The magnitude of kinetic isotope effects in solvolysis may be distorted by possible multiple reaction pathways such as competing elimination, participation, ion pairing, rearrangements and other factors which might complicate unequivocal interpretation. Equilibrium isotope effect measurements in

ISOTOPE EFFECTS

ON

153

NMR SPECTRA

degenerate rearrangements of persistent carbocations involve only the properties of well-defined species and can be evaluated very accurately by nmr spectroscopy. The 2,3-dimethylbicyclo[2.2.2]octyl cation [ 1521 and the methyl-substituted 2-heptyl cations [95] and [97] are suitable models for different orientations of a p-C-D bond with respect to the vacant p-orbital at the C+-carbon.The cations [ 1521, [95] and [97]undergo fast 1,2-and 1,5-hydride shifts (105)-( l07), which give averaged nmr-signals and establish an isotope exchange situation for p-H vs p-D. Both prerequisites for evaluation of the conformational dependence by the isotopic perturbation method are fulfilled.

&CH3

(105)

c

H3C

/+\

CH3H

p-CH3

\

CH,

2,3-Dimethylbicyclo/2.2.2 Joctyl cation. The 3C nmr spectrum of cation [I 531 shows typical splittings of the averaged peaks C-l/C-4, C-5/C-6 and C-7/C-8 due to the lifting of the degeneracy of the equilibrium (108) (Siehl and Walter, 1984). The signal for the averaged C-2/C-3 carbons is merged

154

HANS-ULLRICH SIEHL

into the baseline due to the very large kinetic line broadening (AG* = 4.7 kcal mol-' at - 122°C). The isotope splitting for the CHJCD, groups is 6.4 ppm ( - 122°C) and not symmetrical with respect to the protio-cation [152]. The CD,-group was found shifted upfield and has an additional intrinsic upfield shift of 0.9ppm. This shows that cation [153b]

with the CD, group further removed from the positive charge is favoured in equilibrium (108). The size of the equilibrium constant is KCD3= I .73 at - I20'-C. The isotope splittings and the equilibrium constant are strongly temperature-dependent as expected. The thermodynamic parameters A H o = -65calmol-' per D and A 9 = -0.06calmol-'K-' per D were determined from the temperature dependence of K between - 137°C and -92°C. These values are in good agreement with those reported for other P-methyl-deuteriated carbocations.

In marked contrast, the bridgehead deuteriated cation [I541 did not show any sizeable isotope splittings, except small temperature-independent intrinsic isotope shifts. The broadest peak, i.e. the averaged C,/C, peak, has a line width of 42 Hz at - 122°C. which gives a calculated maximum for K of 1.03 per D at - 122 C. The absence of equilibrium isotope splittings in [I541 was confirmed by the ' H nmr spectrum, which is virtually identical with that of the unlabelled cation [I521 except that the peak for the bridgehead protons has only half the intensity. These results have been interpreted as a direct proof of the hyperconjugational origin of the dependence of the P-deuterium equilibrium isotope effect on dihedral angle in a persistent carbocation. In [I531 maximum overlap between the vacant p-orbital and one of the methyl C-D or C-H bonds is always possible leading to large equilibrium isotope splittings. In [I541 the dihedral angle of 90" permits no overlap of the vacant p-orbital with the


155

bridgehead C-H and C-D bonds. Hyperconjugation with the bridgehead protons or deuteriums is sterically suppressed. In the parlance of valence bond theory, hyperconjugation involving the bridgehead position is inhibited because the “no bond” resonance structure would be a bridgehead olefin and thus too unstable to contribute significantly to the hyperconjugational stabilisation of the cation (Siehl and Walter, 1984). 2,6-Dirnethyl- and 2.4.4,6-tetrameth~~l-heptyl cations The total isotope effect of a CD,-group is independent of conformation (Sunko et al., 1977), and the CD,-substituted cations [I551 and [I561 can therefore serve as a model for a dihedral angle of 0” whereas the cations [ 1571 and [ 1581 deuteriated in the methylene group provide information on the conformational dependence of the equilibrium isotope effect.

H2C I

I L

CH2

I

/+\ c

,?,-CD CH3 CH,

H3C

H3c-c H3C

c

/*\ H3C CD,

c H,C

/+\

( II l l )

3

/CwCD3 CIY 3 H \CH3

[156]

The methyl groups on C-2/C-6 and the C-3/C-5 methylene groups give averaged peaks in the ‘H and 13C nmr spectra of cations [95] and [97] (p. 153). The averaged peak for the C’/C-H carbons is not visible due to large kinetic line-broadening caused by the fast 2,6-hydride shift (Saunders and Siehl, 1985; Siehl and Walter, 1985). The p-methyl deuteriated cations [I 551 and [ 1561 experience large temperature-dependent isotope splittings for the averaged peaks. As expected the deuteriated methyl groups prefer the position remote from the positive charge.

.

156

HANS-ULLRICH SIEHL

Equilibrium constants for [ 1551, K(CD,) = 1.69 at - 120°C, and for [ 1561, K(CD,) = 1.91 at - 123"C, were obtained from the isotope splittings of the averaged CH,- and CH,-groups in 'H and 13C nmr spectra. Thermodynamic parameters for [ 1551 are AW = - 62.4 cal mol- ', ASO = - 0.06 cal mol-' K - ' , and for [156] AH" = -65.7 cal mol-', AS" = -0.01 cal mol-' K - ' per deuterium, determined from the 13C spectra.

H3Cc' H3c/'H

3

c

1:

r L

/+\

H3C

CH3

,-, c

3

c,

3c\_

3

C FC iH3 H3C/+\ CH3 H \ CH3 [ I581

The methylene-deuteriated cations [ 1571 also showed large temperaturedependent isotope splittings resulting from pertubation by one (K = 1.31 at - 110°C) and two deuteriums (K = 1.72 at - 110°C). The cations with deuterium remote from the positive charge are favoured in (1 12) equilibrium by AHo = -125calmol-' (dl-[157]) and AW = -129calmol-' per D (D2-[1571) respectively. No splittings appear in the spectra of the methylene-deuteriated cations [ 1581 and only small temperature-independent intrinsic isotope shifts are observed in the 13C nmr spectrum. The kinetically broadened signal of the averaged methyl groups in D,-[ 1581 has a line width of 118 Hz at - 130°C which gives a calculated maximum for K of 1.06 per D. The proton spectra are identical with the spectra of the unlabelled cation [97] except for the reduced intensity of the partially deuteriated methylene groups. Different conformational preferences in [ 1571 and [ 1581 result in suppression of the equilibrium isotope effect in [158]. Two limiting conformations [159] and [160] for the cations [157] and [158] can be envisaged.


157

Conformation [ 1601 has a dihedral angle of 30" between the empty p-orbital and the b-methylene C-D/C-H bonds and should show a large isotope effect, whereas conformation [I591 with an angle of 60" should show a smaller effect (Sunko et al., 1977; Hehre et al., 1979). The conformational equilibrium between [I591 and [160] is dependent on the size of R. In cation [ 1571 (R = CH,CH,CHMe,) the steric hindrance between R and the vicinal Me groups is less than in [158] where R is more sterically demanding (R = CMe,CH,CHMe,). The large equilibrium isotope effect observed in [157] shows that the preferred conformation for this cation is [160] in which the P-C-H/C-D bonds are better aligned for hyperconjugation with the vacant p-orbital at the C + carbon than in [159]. The severe crowding prevents cation [ 1581 from adopting conformation [ 1601 and conformation [159] is preferred instead. The FC-H/C-D methylene bonds are at unfavourable angles for hyperconjugation in [ 1591 and thus the equilibrium isotope effect is suppressed.

H3:q.j$3...c45. R [IS')]

[ I601

The cationic transition states in the solvolysis of [150] and [ 15 13 are closely related static analogues of cations [157] and [I 581. They prefer conformation [ 1601 (R = CH,CH,) and [ 1591(R = C(CH,),) respectively. This shows that the same type of geometric requirements for C-H hyperconjugation apply to persistent carbocations as was concluded by Shiner for the transition state in solvolytic substitution reactions. The similarity between the equilibrium isotope effects in [157] and the 1,2-dimethylcyclopentyl cation [ 1041 points to a similar conformation in the vicinity of the C+-carbon. The magnitude of the equilibrium isotope effect can thus be used as a new means of aiding in the assignment of conformation and structure in solution. The other major factor which influences the extent of hyperconjugation besides conformational preferences is the degree of charge at the cationic carbon. The demand for C-H hyperconjugation varies according to what other modes of stabilisation are available, and this depends on the structure of the cation especially the substitution pattern at the cationic carbon. Variable demand for C-H hyperconjugation will in turn influence the magnitude of the P-deuterium equilibrium isotope effect. This is clearly evident comparing the P-CD,-isotope effect in 1,2-dimethylcyclopentyl

158

HANS-ULLRICH SIEHL

cation [ I 151 with the decreasing effect with increasing o-bridging in dimethylbicyclo[’. I . Ilhexyl cation [ 1 161 and dimethylnorbornyl cation [ 1 171 (p. 132). C-C-Hyperconjugation in carbocations with suitable structure and appropriate aligned conformations reduces the demand for C-H hyperconjugation and thus changes the magnitude of the P-deuterium isotope effect. The influence o n vibrational frequencies of P-CH, and P-CH,-groups could be different. Depending on these influences and on the preferred pathway (interior-CH, vs terminal-CH,) for C-H hyperconjugation, discussed by Saunders et al. ( 1977a), the equilibrium isotope effect for P-methylene deuteriation can be larger, for example in [I041compared to [116] (p. 149), or smaller, for example in [ 1581 compared to [ 1561 (p. 155) (Siehl, I986b), than predicted from the corresponding CD,-effects using a cos’ function (Sunko and Hehre, 1983). References

P., Engdahl. C. and JonsCI1. G (1981). J. Am. Chem. SOC.103, 1583 P., Jonsall, G . and Engdahl, C. (1983a). Adv. Phys. Org. Chem. 19, 223 P., Engdahl, C. and Jonsall, G. (1983b). J . Am. Chem. Soc. 105, 891 P., Jonsall, G., Huang, M. B. and Goscinski, 0. (1983~).J . Chem. Soc. Perkin Trans 2 305 Anet. F. A. L. and Dekmezian, A. H. (1979). J . Am. Chem. Soc. 101, 5449 Anet. F. A. L. and Kopelevich. M. (1986a). J . Am. Chem. Soc. 108, 1355 Anet, F. A. L. and Kopelevich. M. (1986b). J . Am. Chem. Soc. 108, 2109 Anet. F. A. L. and Leyendecker, F. (1973). J . Am. Chem. Soc. 95, 156 Anet. F. A. L. and Rawdah. T. N. (1978). J . Am. Chem. Soc. 100, 7166 Anet, F. A. L., Cheng, A. K . and Krane, J. (1973). J . Am. Chem. Soc. 95, 7877 Anet. F. A. L. Cheng. A. K.. Mioduski. J. and Meinwald, J . (1974). J . Am. Chem. Soc. 96, 2887 Askani, R., Kalinowski, H.-0. and Weuste, B. (1982). Org. Mag. Res. 18, 176 Askani, R., Kalinowski, H.-O., Pelech, B. and Weuste, B. (1984). Tetrahedron Lett. 25, 2321 Aydin. R. and Gunther. H . (1981). J . Am. Chem. Soc. 103, 1301 Batiz-Hernandez, H . and Bernheirn. R. A. (1967). f r o g . N M R Specrrosc. 3, 63 Berger. S. and Diehl, B. W. K. (1986). Mug. Res. Chem. 24, 1073 Berger. S. and Kunzer. H . (1985). J . Am. Chem. Soc. 107, 2804 Bigeleisen. J. and Mayer, M. G. (1947). J . Chem. Phys. 15, 261 Booth, H . and Everett, J. R. (1980a). Can. J . Chem. 58, 2709 Booth, H. and Everett, J. R . (1980b). Can. J . Chem. 58. 2714 Brookhart, M.. Lamanna, W. and Humphrey, M. B. (1982). J . Am. Chern. SOC.104, 21 I7 Brown, H. C. (1983). Ace. Chem. Res. 16, 432 Brown. H . C. and Schleyer, P. v. R. (1977). In “The nonclassical ion problem”, Chap. 5. Plenum Press. New York Bywater. S., Brownstein, S . and Worsfold, D. J . (1980) J . Organomet. Chem. 199, 1 Casey, C. P.. Fagan, P. J. and Miles. W. H . (1982). J . Am. Chem. SOC.104, I134

Ahlberg. Ahlberg, Ahlberg, Ahlberg,


159

Chan, S. I.. Lin, L.. Clutter, D. and Dea. P. (1970). Proc. Nutl. Acad. Sci. U.S.A.65, 816 Chrisll, M., Leininger, H. and Bruckner, D. (1983). J . Am. Cheni. Soc. 105. 4843 Coates, R. M. and Fretz. E. R. (1975). J . Am. Chem. SOC.97. 2538 Collins, C. J. and Bowman. N. S. (eds.) (1970). “Isotope Effects in Chemical Reactions”, Van Nostrand Reinhold, New York DePuy, C. H., Fiinfschilling, P. C. and Olson, J. M. (1976). J . Am. Chem. SOC.98, 276 DePuy, C. H., Funfschilling, P. C., Andrist. A. H. and Olson, J. M. (1977). J. Am. Chem. SOC.99, 6297 Dewar, M. J. S. and Merz, K . M. Jr. (1986). J . Am. Chem. SOC.108, 5634 Dewar, M. J . S. and Reynolds, C. H . (1984). J . Am. Chem. Soc. 106. 6388 Ernst, L., Hopf, H. and Wullbrandt. D. (1983). J . Am. Chem. SOC.105, 4469 Fong. F. K . (1974). J . Am. Chem. SOC.96, 7638 Forsen, S.. Gunnarsson, G.. Wennerstrom, H., Altman, L. J. and Laungani, D. (1978). J . Am. Cheni. SOC. 100, 8264 Forsyth, D. A. (1984). In “Isotopes in Organic Chemistry” (eds. E. Buncel and C. C. Lee), Vol. 6, Chap. I . Elsevier, Amsterdam Forsyth, D. A. and MacConnell, M. M. (1983). J . Am. Chem. SOC.105, 5920 Forsyth. D. A. and Pan, Y. (1985). Tetrahedron Lett. 26, 4997 Forsyth, D. A. and Yang. J.-R. (1986). J . Am. Chem. SOC.108, 2157 Forsyth, D. A,, Botkin, J. H. and Osterman. V. M. (1984). J . Am. Chem. SOC.106, 7663 Forsyth. D. A. Botkin, J. H. and Sardella, D . J. (1986). J . Am. Chem. SOC.108,2797 Gold, V. (1968). Truns. Farads), SOC. 64, 2770 Grob, C. A. (1983). A N . Chem.. Res. 16, 426 Grubbs, R. H.. Lee, B. J., Gajda, G . J., Schaefer, W. P., Howard, T. R.. Ikariya, T. and Straus, D. A. (1981). J . Am. C h ~ mSOC. . 103, 7358 Gunther, H. and Aydin, R. (1981). Angew. Chem. 93, 1000 Gunther, H. and Wesener. J. R. (1982). Tetrahedron Lett. 23, 2845 Gunther, H . and Wesener, J. R. (1985). J. Am. Chem. SOC.107, 7307 Gunther, H.. Pawliczek. J.-B., Ulsen, J. and Grimme, W. (1975). Chem. Ber. 108, 3141 Gunther, H., Joseph-Natan, P., Aydin, R., Wesener, J. R., Santillan, R. L. and Garibay, M. E. (1984). J . Org. Chem. 49, 3847 Giinther, H.. Wesener, J. R. and Moskau, D. (1985). Tetrahedron Left. 1491 Halevi, E. A . (1963). Prog. Phys. Org. Chem. 1, 109 Halevi, E. A., Bary, Y. and Gilboa, H. (1979). J . Chem. SOC.Perkin 2 938 Hansen, H.-J. and Lang, R. W. (1980). Helv. Chim. Acfa 63, Fasc. 5, 1215 Hansen. P. E. (1983). Ann. Rep. on N M R Spectra. 15, 105 Hansen, P. E. and Duus, F. (1984). Org. Mag. Res. 22, 16 Hansen, P. E. and Lyfka, A. (1984). Org. Mag. Res. 22, 569 Hansen, P. E., Duus, F. and Schmitt, P. (1982). Org. Mag. Res. 18, 58 Hehre, W. J., DeFrees, D. J. and Sunko, D. E. (1979). J . Am. Chem. SOC.101,2323 Hehre, W. J., Hout, Jr, R. F. and Levi, B. A. (1983). J . Compufarional Chem. 4, 449 Hogeveen, H. and van Kruchten, E. M. G . A. (1981). J . Org. Chem. 46, 1350 Ittel, S. D., Van-Catledge, F. A. and Jesson, J. P. (1979). J . Am. Chem. SOC.101,6905 Jackman, L. M. and Cotton, F. A. (eds) (1975). “Dynamic Nuclear Magnetic Resonance Spectroscopy”. Academic Press, New York Jamcson, C. J. (1977). J . C h m . Phys. 66, 4977

160

HANS-ULLRICH SlEHL

Jameson. C. J. and Osten, H.-J. (1984a). J . Chem. Phys. 81, 4293 Jameson, C. J. and Osten. H.-J. (1984b). J . Chem. Phys. 81, 4300 Jensen. F. R. and Neese, R. A. (1971). J . Am. Chem. SOC.93, 6329 Jensen, F. R. and Smith, L. A. (1964). J . Am. Chem. SOC.86, 956 Jeremik D., Milosavljevic, S. and Mihailovic, M. L. (1982). Tetrahedron 38, 3328 Kalinowski, H.-0. (1984). Nuchr. Chem. Tech. Lab. 32, 874 Kalinowski, H.-0.. Berger, S. and Braun, S. (1984). “13 C-NMR Spektroskopie”. Thieme Verlag, Stuttgart Karle, J., Lowrey, A. H., George, C. and D’Antonio, P. (1971). J . Am. Chem. SOC. 93, 6399 Kates, M. R. (1978). Ph.D. Thesis. Yale University, New Haven, Connecticut Koch, E.-W. (1985). Dissertation. University of Tubingen, Tubingen, Germany Krcsgc. A. J., Arrowsmith. C. H., Baltzer, L., Powell, M. F. and Tang, Y. S. (1986). J . A m . Chem. SOC.108, 1356 Leipert, T. K . (1977). Org. Mug. Res. 9, 157 Lippmaa, E., Pehk, T. and Laht, A. (1982). Org. Mug. Res. 19, 21 Lloyd, J. R. (1978). Ph.D. Thesis. Yale University, New Haven, Connecticut Maier. G., Kalinowski, H.-0. and Euler, K. (1982). Angew. Chem. 94, 706 (Int. Ed. Engl., 21, 693) Majerski, Z., Zuanic, M. and Metelco, B. (1985). J . Am. Chem. Soc. 107, 1727 McMurry, J. E. and Hodge, C. N. (1984). J . Am. Chem. SOC.106, 6450 Melander, L. (1960). “Isotope Effects on Reaction Rates”. Ronald Press Company, New York Melander, L. and Saunders, Jr. W. H. (1980). “Reaction Rates of Isotopic Molecules”. John Wiley, New York Moore, P., Howarth, 0. W., McAteer. C. H. and Morris, G. E. (1981). J . Chem. SOC. Chem. Cornrnun. 506 Myhre, P. C., Yannoni, C. S., Kruger, J. D., Hammond, B. L., Lok, S. M., Macho, V., Limbach, H. H. and Vieth, H. M. (1984). J . Am. Chem. SOC. 106, 6079 Oki, M. (1985). “Applications of Dynamic N M R Spectroscopy to Organic Chemistry”. VCH Publishers, Weinheim, Germany Olah, G. A,, Kelly, D. P., Jeuell, C. L. and Porter, R. D. (1970). J . Am. Chem. Soc. 92, 2544 Olah, G. A., Jeuell, C. L., Kelly, D. P. and Porter, R. D. (1972). J . Am. Chem. SOC. 94, 146 Olah. G. A,, Prakash, G. K. S., Donovan, D. J. and Yavari, I . (1978a). J . Am. Chem. Soc. 100, 7085 Olah. G . A,, Roberts, J. D., Staral, J . S., Yavari, I., Prakash, G . K. S. and Donovan, D. J. (1978b). J . Am. Chem. SOC.100, 8016 Olah, G. A., Prakash, G. K. S. and Nakajima, T. (1982). J . Am. Chem. SOC.104,1031 Olah, G . A,, Saunders, M. and Prakash, G. K. S. (1983). Acc. Chem. Res. 16, 440 Olah, G. A,, Krishnamurthy, V. V., Prakash, G. K. S. and Iyer, P. S. (1985a). J . Am. Chem. SOC.107, 5015 Olah, G. A,, Prakash, G. K. S. and Arvanaghi, M. (1985b). J . Am. Chem. SOC.107, 6017 Olah. G. A,. Sommer, J. and Prakash. G. K. S. (1985~).In “Superacids”, p. 144. John Wiley. New York Petersen, S. B. and Led, J. J. (1979). J . Mug. Res. 3, 603 Rappoport, Z . , Biali, S. E. and Hull, W. E. (1985). J . Am. Chem. Soc. 107, 5450 Reuben, J. (1984). J . Am. Chem. SOC.106, 6180


161

Reuben, J. (1985a). J . Am. Chem. SOC.107, 1756 Reuben, J. (1985b). J . Am. Chem. SOC.107, 5867 Reuben, J. (1985~).J . Am. Chem. SOC.107, 1433 Reuben, J. (1986a). J . Am. Chem. SOC.108, 1082 Reuben, J. (1986b). J . Am. Chem. SOC.108, 1735 Roberts, J. D., Brittain, W. J. and Squillacote, M. E. (1984). J. Am. Chem. SOC.106, 7280 Robinson, M . J. T. (1971). Pure Appl. Chem. 25, 635 Robinson, M. J. T. and Baldry, K. W. (1977a). J . Chem. Res ( S ) 86; J . Chem. Res ( M ) 1001

Robinson, M. J. T. and Baldry, K. W. (1977b) Tetrahedron 33, 1663 Robinson, M. J. T., Rosen, K. M. and Workman, J. D. B. (1977). Tetrahedron 3, 1655 Sandstrom, J. (1982). “Dynamic N M R Spectroscopy”. Academic Press, New York Saunders, M. (1979). In “Stereodynamics of Molecular Systems” (ed. R. Sarma), p. 171. Pergamon Press, Oxford Saunders, M. and Evilia, R. F. (1985). Spectrosc. Lett. 18, 105 Saunders, M. and Handler, A. (1985). Unpublished results; c j Saunders et al. Abstr. “Symposium on Advances in Carbocation Chemistry”, American Chemical Society. Seattle meeting, March 1983 Saunders, M. and Jarret, R. M . (1985). J . Am. Chem. SOC.107, 2648 Saunders, M . and Jarret. R. M . (1986). J . Am. Chem. SOC.108, 7549 Saunders, M. and Kates, M. R. (1977). J . Am. Chem. SOC.99, 8071 Saunders, M . and Kates, M. R. (1978). J . Am. Chem. SOC.100, 7082 Saunders, M . and Kates, M. R. (1980). J . Am. Chem. Soc. 102, 6867 Saunders, M. and Kates, M. R. (1983). J . Am. Chem. SOC.105, 3571 Saunders, M. and Siehl. H.-U. (1980). J . Am. Chem. SOC.102, 6868 Saunders, M . and Siehl, H.-U.‘ (1985). Unpublished results; cb Siehl and Walter (\1 985), Saunders, M. and Vogel, P. (1971a). J . Am. Chem. SOC.93, 2559 Saunders, M . and Vogel, P. (1971b). J . Am. Chem. SOC.93, 2561 Saunders, M., Vogel, P. and Jaffe, M. H. (1971). J. Am. Chem. SOC.93, 2558 Saunders, M., Wiberg, K. B., Seybold, G. and Vogel, P. (1973). J . Am. Chem. SOC. 95. 2045 Saunders. M., Telkowski, L. and Kates, M. R. (1977a). J . Am. Chem. SOC.99, 8070 Saunders, M., Wiberg, K. B., Kates, M. R. and Pratt, W. (1977b). J . Am. Chem. SOC. 99, 8072 Saunders, M., Anet, F. A. L., Hewett, A. P. W. and Basus, V. J. (1980a). J . Am. Chem. SOC.102, 3945 Saunders, M . , Faller, J. W. and Murray, H. H. (1980b). J . Am. Chem. SOC.102,2306 Saunders, M., Schleyer, P. v. R. and Chandrasekhar, J. (1980~).Rearrangements of carbocations. In “Rearrangements in Ground and Excited States” (ed. P. de Mayo), Chap. 1. Academic Press, New York Saunders, M., Walker, G. E. and Kates, M. R. (1981). J. Am. Chem. SOC.103, 4623 Saunders, M., Saunders, S. and Johnson, C . A. (1984). J . Am. Chem. SOC.106,3098 Schleyer, P. v. R. and Neugebauer, W. (1980). J . Organometal. Chem. 198, CI Schleyer, P. v. R., Baborack, J. C. and Chari, S. (1971). J . Am. Chem. SOC.93, 5275 Schlosser. M. and Strahle, M. (1980). Angew. Chem. 92, 497 (Int. Ed. Engl. 19,487) Schlosser, M. and Strahle, M . (1981) J . Organometal. Chem. 220, 277 Schneider, J. ( 1 985). Diplomarbeit. University of Tiibingen, Tiibingen, Germany

162

HANS-ULLRICH SlEHL

Schrock, R. R. and Holmes. S. J. (1981). J . Am. Chem. SOC.103,4599 Servis. K. L. and Domenick, R. L. (1985). J. Am. Chem. Sac. 107, 7186 Servis, K. L. and Domenick. R. L. (1986). J . Am. Chem. SOC.108, 2211 Servis. K. L. and Shue. F.-F. (1980). J . Am. Chem. SOC.102, 7233 Shapet’ko, N . N., Berestova. S. S., Lukovkin, G . M . and Bogachev, Y. S. (1975). Org. Mug. Res. 7, 237 Shapiro. M . J., Kahle. A. D., Damon, R. and Coppola, G . M. (1981). J. Org. Chem. 46. 1221 Shapley, J. R. and Calvert, R. B. (1977). J. Am. Chem. SOC.99, 5225 Shapley. J. R. and Calvert, R. B. (1978). J . A m . Chem. SOC.100, 7726 Shapley, J. R., Williams. J. M., Stucky. G . D., Calvert, R. B., Schultz, A. J. and Suib, S. L. (1978). J . Am. Chem. SOC.100, 6240 Shapley. J. R., Williams, J. M., Stucky, G. D., Calvert, R. B., and Schultz, A. J. (1979). horn. Chem. 18, 319 Shapley, J. R.,Cree-Uchiyama, M . and George, G. M . St. (1986). J. Am. Chem. SOC. 108. 1316 Shiner, Jr., V. J. (1961). J . A m . Chem. SOC.83, 240 Shiner, Jr., V. J. and Hartshorn, S. R. (1972). J . Am. Chem. SOC.94. 9002 Shiner, Jr., V. J. and Humphrey, Jr., J. S. (1963). J. Am. Chem. SOC.85, 2416 Siehl, H.-U. (1985). J. Am. Chem. SOC.107, 3390 Siehl, H.-U. (1986a). Proc. 8th IUPAC Conference on Physical Organic Chemistry, Tokyo, Japan, Aug. 1986, Elsevier Science Publishers, Amsterdam, in press Siehl, H.-U. (1986b). Habilitationsschrift. University of Tubingen, Tubingen, Germany Siehl, H.-U. and Koch, E.-W. (1985). J. Chem. SOC.Chem. Commun. 496 Siehl, H.-U. and Walter, H. (1984). J . Am. Chem. SOC.106, 5355 Siehl, H.-U. and Walter, H. (1985). J . Chem. SOC.Chem. Commun. 76 Sorensen, T. S. and Kirchen, R. P. (1977). J . Am. Chem. SOC.99, 6687 Sorensen, T. S. and Kirchen, R. P. (1978). J. Chem. SOC.Chem. Commun. 769 Sorensen, T. S. and Kirchen, R. P. (1979). J . Am. Chem. SOC.101, 3240 Sorensen, T. S. and Ranganayakulu, K. (1984). J. Org. Chem. 49,4310 Sorensen, T. S. and Schmitz, L. R. (1980). J. Am. Chem. SOC.102, 1645 Sorensen, T. S., Ranganayakulu, K. and Huang, E. (1972). J . Am. Chem. SOC.94, 1780 Sorensen, T. S., Haseltine, R. and Wong, N. (1975). Can. J . Chem. 53, 1891 Sorensen, T. S., Kirchen, R. P. and Wagstaff, K. (1978). J. Am. Chem. SOC.100,6761 Sorensen, T. S., Kirchen, R. P., Ranganayakulu, K., Rauk, A. and Singh, B. P. (1981a). J . Am. Chem. SOC.103, 588 Sorensen, T. S., Kirchen, R. P., Okazawa, N., Ranganayakulu, K. and Rauk, A. (1981b). J . Am. Chem. SOC.103, 597 Sorensen, T. S., Kirchen, R. P., Ranganayakulu, K. and Singh, B. P. (1981~).Can. J . Chem. 59, 2 173 Stern, M., Spindel, W. and Monse, E. U . (1965). J. Chem. Phys. 48, 2908 Stone, G . A., Dawkins, G. M., Green, M. and Orpen, A. G . (1982). J . Chem. SOC. Chem. Commun. 41 Sunko, D. E. and Hehre, W. J. (1983). Prog. Phys. Org. Chem. 14, 205 Sunko, D. E., BoriEi?, S. and Nikolertic, M . (1962). J . Am. Chem. SOC.84, 1615 Sunko, D. E., Humski, K., MaloJEiC, R. and BorEiC, S. (1970). J. Am. Chem. SOC.92, 6534 Sunko, D. E., Szele. I. and Hehre. W. J. (1977). J . Am. Chem. SOC.99, 5000


163

Swain. C. G., Stivers. E. C., Reuver, J. F. and Schaad, L. J. (1958). J . Am. Chem. SOC.80, 5885 Telkowski, L. (1975). Ph.D Thesis. Yale University, New Haven, Connecticut Vogel, P. (1979). Carbokationen. In Wentrup, C. “Reaktive Zwischenstufen” (ed. C. Wentrup), Vol. 2, Chap. 7. Thieme Verlag, Stuttgart Vogel, P. (1985). “Carbocation Chemistry”, Elsevier Science Publishers, Amsterdam Walling, C. (1983). Acc. Chem. Res. 16,448 Walter, H. (1 985). Dissertation. University of Tubingen, Tubingen, Germany Wehrli, F. W. and Wirthlin, T. (1976). “Interpretation of Carbon-I3 N M R Spectra”, p. 108. Heyden, London Wehrli, F. W., Jeremic, D., Milhailovik, M. L. and Milosavljevic, S. (1978). J . Chem. SOC.Chem. Commun. 302 Whipple, E. B., Andrews, G . C. and Chmurny, G.N. (1981). Org. Mag. Res. 15,324 Willi, A. V. (1983). “Isotopeneffekte bei chemischen Reaktionen”. Thieme Verlag, Stuttgart Williams, I. H.(1986). J . Chem. SOC.Chem. Commun. 629 Williams, J . M.,Ittel, S. D., Brown, R. K., Schultz, A. J., Stucky, G. D. and Harlow, R. L. (1980). J . Am. Chem. SOC.102, 981 Wolfsberg, M. (1969). Ann. Rev. Phys. Chem. 20,449 Wolfsberg, M.(1972). Acc. Chem. Res.5, 225 Wolfsberg. M. and Kleinman, L.I. (1973). J . Chem. Phys. 59, 2043 Wolfsberg. M. and Kleinman. L. I. (1974a). J . Chem. Phys. 60, 4740 Wolfsberg, M. and Kleinman, L. I. (1974b). J . Chem. Phys. 60, 4749 Yannoni. C. S.,Myhre, P. C. and Macho, V. (1982). J . Am. Chem. SOC.104, 7380 Yannoni, C.S.,Myhre, P. C. and McLaren, K. L. (1985). J . Am. Chem. SOC.107, 5294 Yashiro, M., Yano, S. and Yoshikawa, S. (1986). J . Am. Chem. SOC.108, 1096


The Mechanisms of Reactions of p-Lactam Antibiotics M I C H A E L I. P A G E Department of Chemical and Physical Sciences, The Polytechnic, Queensgare, Huddersfield HD1 3DH. U.K 1 Introduction 166 2 Mode of action of j3-lactam antibiotics 173 3 Is the antibiotic b-lactam unusual? 184 Structural and ground-state effects 186 Kinetic effects 194 Summary 198 4 Alkaline hydrolysis and structure: chemical reactivity relationships Penicillins 199 Cephalosporins 202 Acid hydrolysis 207 Spontaneous hydrolysis 2 15 Buffer catalysed hydrolysis 216 Metal-ion catalysed hydrolysis 218 Coordination site 219 Rate enhancement 221 Effect of transition metal ion 222 Effect of p-lactam 222 9 Micelle catalysed hydrolysis of penicillins 223 10 Cycloheptaamylose catalysed hydrolysis 232 1 1 The aminolysis of p-lactam antibiotics 233 Intermolecular general base catalysis 234 Intramolecular general base catalysis 241 Intramolecular general acid. catalysis and the direction of nucleophilic attack 243 Uncatalysed aminolysis 244 Metal-ion catalysed aminolysis 246 Imidazole catalysed isomerisation of penicillins 248 Intramolecular aminolysis 249 12 The stepwise mechanism for expulsion of C(3’)-leaving groups in cephalosporins 250 ADVANCES I N PHYSICAL ORGANIC CHEMISTRY VOLUME 23 I S B N 0 I2 033523 9

198

Copyrrghr Q 1987 Academic Press Inc. London, All riRhls of reproduction in any form resewed.

165

MICHAEL I. PAGE

166

Reaction with alcohols and other oxygen nucleophiles 252 Thiazolidine ring opening 255 14 Epimerisation of penicillin derivatives 258 References 261 13

1

Introduction

Nearly sixty years have now passed since Sir Alexander Fleming observed antibiosis between a Penicillium mould and bacterial cultures and gave the name penicillin to the active principle (Fleming, 1929). Although it was proposed in 1943 that penicillin contained a p-lactam ring [I] (Abraham et al., 1943), this was not generally accepted until an X-ray crystallographic determination of the structure had been completed (Crowfoot et al., 1949). Penicillin was the first naturally occurring antibiotic to be characterized and used in clinical medicine. It can now be seen as the progenitor of the p-lactam family of antibiotics which are characterized by the possession of the four-membered p-lactam ring. Penicillins were originally detected in fungi but subsequently were found in streptomycetes (Miller et al., 1962). The cephalosporins 121, the second member of the p-lactam antibiotic family, were also originally discovered in fungi but later detected in streptomycetes (Newton and Abraham, 1955; Higgins et al., 1974). Until 1970 penicillins and cephalosporins were the only examples of naturally occurring p-lactam antibiotics. The discovery of 7-a-methoxycephalosporins [3] from Streptomyces in 1971 (Nagarajan et al., 1971) stimulated the search for novel p-lactam antibiotics of microbial origin using sensitive new screening procedures. At present, p-lactam antibiotics can be classified into several groups according to their structure: 1

Penicillins (penams)

2

Cephalosporins (cephems)

3 Cephamycins 4 Oxacephems

5 Penems 6 Clavulanic acid 7 Thienamycin (carbapenems) 8 Nocardicins 9 Monobactams

p-lactams fused to a thiazolidine. p-lactams fused to a dihydrothiazine. 7-a-methoxycephalosporins. replacement of cephalosporin ring S by 0. double bond in 5-membered thiazolidine ring. oxapenam derivatives, 0-lactam fused to an oxazolidine. no heteroatom in the ring fused to the p-lactam. monocyclic 0-lactams. monocyclic 0-lactams of sulphamic acid.

MECHANISMS OF REACTIONS OF p-LACTAM ANTIBIOTICS

167

RCONH

0

COOH

COOH

PI

[31

[41

RCONH

0 F x I 0 2 H [51

[61

RCONH

7-1

168

MICHAEL I. PAGE

CONH Monobactam Nocardicin R C O N H p N p o H 0

\

COOH

\

Penem

R

C

O

N

Osacephem

H ' j ~j

-

R 1

-COOH

0

/

\ Clavulanic ac

RcoN:n>R, 0

Cephamycin

COOH I

COOH

Cephslosporin

RcoN%q; 0

COOH

Penicillin

Xscowrcd 192N

FIG. 1 The discovery of p-lactam antibiotics displayed against Anno Domini

MECHANISMS OF REACTIONS OF 8-LACTAM ANTIBIOTICS

169

Penam 4-Thia- I -azabicyclo[3.2.0]heptan-7-one

A’-Penem 4-Thia- 1 -azabicyclo[3.2.0]hept-2-en-7-one

Clavam or Oxapenam 4-Oxa-1 -azabicyclo[3.2.0]heptan-7-one

rn

0

Carbapenem 1-Azabicyclo[3.20]hept-2-en-7-one

Monobactam Azetidin-2-one

1:’:: N 4

Ce p ham

0

A’-Cephem 0

;1!3

A 3-Cephem

0

FIG. 2

Basic structural units of common p-lactam antibiotics

The chronology of these antibiotics is illustrated in Fig. 1, and some trivial names used to describe the ring structures are shown in Fig. 2 . P-Lactams have now been found in eukaryotic fungi, actinomycetes (mostly in the genus Streptomyces) and even in bacteria, but their in vivo role is not yet understood. Recently p-lactones have also been isolated from various micro-organisms (Sykes et al., 1985).

170

MICHAEL I. PAGE

A large number of nuclear analogues of the p-lactam antibiotics have been prepared by complete chemical synthesis or by partial synthesis starting from a naturally occurring p-lactam. Of these only the oxacephems [4] and the penems [5] appear to have been investigated clinically. Most semisynthetic penicillins are made by the acylation of 6-aminopenicillanic acid [ 101 whilst cephalosporins can be prepared to give a variety of side chain substituents at C(7) and C(3) [ l I]. The latter illustrates a problem with nomenclature. Classically P-lactam antibiotics have been described using the ring sulphur as position-1. Hence for penicillins and cephalosporins the numbering is as illustrated in [12] and [I I], respectively. Similarly substituents in these bicyclic systems are usually described by the prefix a- or Prather than the equivalent exo or endo description. RCONH N 0

CO,H

CO,H

C0,H [I21

6-P-Side chains of penicillins may confer stability to acids and have allowed the development of orally administered penicillins, [ 121. These derivatives are unaffected by gastric juice and may be absorbed from the gut. In contrast, although most cephalosporins are intrinsically stable at the low pH of gastric juice, the number of 7-P-side chains allowing oral absorption is limited. Some examples of oral cephalosporins are given in Table 1. The so-called first and second generation injectable cephalosporins are 3,7modified cephalosporins (Table 1). Although 7-P-side chains may improve potency and breadth of antibacterial activity their potential is only expressed when the molecule also contains an appropriate 3-side chain. Third generation cephalosporins differ in their increased spectrum of activity, potency and high cost: “And they did worship strange Lactams, even unto the third generation” (Berger, 1984).

171


TABLE1 Names and structures of common penicillins and cephalosporins

(a) Penicillins RCONH

Lcx

0

-

CO,H Penicillin

R

Benzylpenicillin (Pen G ) PhCHz-

Penicillin Carbenicillin

Pen F Pen X

Pen K

Pen V Pen N

Propicillin

Phenbenicillin

PhOCH(Pht

Diphenicillin Ph

R PhCH-

I

TABLE1 (continued) Penicillin

R

R

Penicillin OMe

/

Ampicillin

PhCH-

Ticarcillin

NHZ

CO,H

(b) Cephalosporins RCONH 0~

c

H

R

Cephalosporin Oral cephalosporins cephaloglycin cephalexin cefaclor cefatrizine

~ CO,H~

L

Ph-CH(N H 2)Ph-CH(NH2)Ph-CH(NH2)HOC6Hd-CH(NHz)-

-0Ac

-H -CI

-scr N” H

First generation

cephalothin

-0Ac

cephaloridine

CHZ-

N=N cefazolin

I

\

N+N-CH,-

z

L

173


Cephalosporin

R

L

Second generation

N-N cefamandole

PhCH(0H)-

AH3 cefuroxime

cephamycin

-0-CONH,

-OCONH2

H02C-CH(CHl)3-

I

NH2 Third generation

/ cefo taxime

cefoperazone

2 Mode of action of fl-lactam antibiotics

To understand how p-lactams kill bacteria requires a little knowledge of how a bacterial cell wall is formed and why bacteria need cell walls. Bacteria, unlike eukaryotic cells, possess a cell wall which is a complex structure in

174

MICHAEL I

PAGE

both Gram-positive and Gram-negative types. Cell walls help to maintain the shape of bacteria whilst the cell membranes are the osmotic barriers that allow the retention of nutrients and the exclusion of other compounds. Gram-positive and Gram-negative bacteria have internal osmotic pressures which are 10 to 30 times and 3 to 5 times, respectively, the external osmotic pressure. The rigid structure of the bacterial wall of both Gram-positive and Gram-negative species is due to the cross linking of linear polysaccharide chains by short segments of peptides called peptidoglycan. It is accepted generally, but not universally (Reinicke et al., 1985), that the biochemical targets for the p-lactams are some of the enzymes concerned with cell-wall synthesis. What is undoubtedly universally accepted is the increasing complexity of drug action as one moves from target enzyme to target cell and from target cell in a culture tube to target pathogen in the complex environment of an infected host. An understanding of the mode of action of any drug requires a knowledge of: the pathway by which the drug reaches its biochemical target, the identity and structure of the biochemical target and the chemistry of how the target is inhibited, 3 the cell’s physiological response to the inhibition of the drug-sensitive reaction, 4 how and why inhibition of this reaction leads to interference of the life cycle of the cell. 1

2

P-Lactams exert their lethal action only on growing bacterial cells. Within a short time of adding penicillin to a bacterial culture, small bulges form in the bacteria often near mid cell where cell division is expected to occur. With time, the bulge enlarges and ultimately the cell membrane ruptures, resulting in death of the cell. These morphological and lytic effects of penicillin are accompanied by the covalent binding of the antibiotic to proteins, called penicillin binding proteins, PBP (Spratt, 1980). The PBPs are located on the inner membrane and, to reach these, the antibiotic has to traverse a variety of chemical and physical barriers (Fig. 3). In Gram-negative bacteria the diffusion of the p-lactam through the outer (plasma) membrane and across the periplasm may be a rate-limiting factor in determining antibacterial effectiveness (Nikaido, 1981). Before detailing the nature of PBPs and their relationship to the mode of action of P-lactam antibiotics, it is worth noting the high toxicity of p-lactam antibiotics against bacteria compared with their low toxicity against eukaryotic cells. The lowest concentration of penicillin capable of inhibiting the growth of a pneumococcal culture is about 6 x 1O-’g per lo8 bacteria per cm‘ (Tomasz. 1983). In contrast the recommended therapeutic dose for carbenicillin is 40 g per day which illustrates the insensitivity of eukaryotic

MECHANISMS OF REACTIONS OF D-LACTAM ANTIBIOTICS

175

6=

I7 - Lac tam antibiotic

Inner membrane

Outer membrane

FIG. 3 The passage of a p-lactam antibiotic to its killing target in Gram-negative bacteria

cells to p-lactam antibiotics, The selective action of these drugs is also compatible with the notion that their action is related to a unique bacterial component, e.g. the cell wall found in prokaryotes but not in the more complex eukaryotes. The bacterial cell wall envelops the cytoplasm and consists of peptidoglycan which is made up of polysaccharide or “glycan” strands that are cross-linked by branched peptide chains. The glycan structure is universal in murein and consists of alternating p- 1,Clinked residues of N-acetyl-D-glucosamine (NAG) and N-acetylmuramic acid (NAM). The glycan is a modified form of chitin and is drawn in the flat ribbon conformation in which chitin is constrained by hydrogen bonds (Tipper, 1970). The short peptide crosslinks are attached to the carboxyl groups on the NAM residues (Fig. 4). The length and nature of the peptide crosslinks vary with bacteria. During biosynthesis of the bacterial cell wall all peptidoglycans carry a pentapeptide which commonly has the sequence: L-ala-D-glu-y-L-X-D-ala-D-ala, in which X is usually an amino carrying residue such as L-lysine or meso-diamino-

MICHAEL I. PAGE

176

c=o I ____ LA

glycan

1

NH I H+H,

peptide

c=o I

NH

C=O

I

CH,

I

CH,

I

CON H-CH -CON H-CH-C02H u-ala

v-ala

I c=o I

-NAG-NAM-N AG-NAM-NAG L-ala

I

v-glu

I

-X--u-ala-X--D-ala

I

X-v-ala-v-ala

I

v-ylu

I

L-ala

I

-NAG-NAM-NAG-NAM-NAG-

FIG. 4 General structure of peptidoglycans


177

pimelic acid. Crosslinking occurs by displacing the terminal D-alanine residue with the free amino group of X on an adjacent pentapeptide, i.e. by a transpeptidation reaction. The actual extent of crosslinking varies with the bacterial species and can be as low as 25% to greater than 80%, but no peptidoglycan is completely crosslinked. In the uncrosslinked sections, the D-ala-D-ala terminal residues may be removed by carboxypeptidase action. The sequence of events involved in the maturing of peptidoglycan during extension of the bacterial cell wall is only partially understood and remains a matter of controversy. The early stages in the biosynthesis of peptidoglycan in which the glycan polymer is uncrosslinked are not penicillin-sensitive. However, the transpeptidation process, which makes the peptidoglycan insoluble, is highly penicillin-sensitive. These reactions are catalysed by a range of peptidases (Ghuysen ef al.. 1985): (i) DD-transpeptidases - catalyse the cross linking of adjacent NAM pentapeptides with loss of the terminal D-alanine residue (Scheme I ) . (ii) DD-carboxypeptidases catalyse the cleavage of the terminal D-alanine from pentapeptide side chains by a hydrolysis reaction. (Scheme I). (iii) endopeptidases - catalyse the hydrolysis of interpeptide links formed by transpeptidation. ~

CHI

+ H,N-CH-CO,H

Scheme 1

There is a basic similarity in the carboxypeptidase and transpeptidase reactions in that in both cases the carbonyl of the penultimate D-alanine is transferred to an exogeneous nucleophile. If the latter is water then hydrolysis results, whereas if it is an amino group of another peptide then aminolysis occurs and the product is a crosslinked dimer of the two peptides. Historically, a plausible model for the molecular basis of the selectivity of p-lactam action was proposed 20 years ago (Tipper and Strominger, 1965). It was suggested that penicillin may be a structural analogue of a conformational isomer of D-alanyl-D-alanine, i.e. the carboxyterminal portion of the cell wall precursor disaccharide-pentapeptide (Fig. 5). It could be noted in

178

MICHAEL I PAGE

c 0’

‘b

FIG. 5 Comparison of the configuration of N-acylated D-alanyl-D-alanine a 3s. 5R. 6R-penicillin ( h )

(0)with

passing that the configuration at C(6) in penicillin is the opposite of that predicted by a complete structural analogy with the D-alanyl residue. Tipper and Strominger also suggested that transpeptidation was a two step reaction involving an acyl-enzyme intermediate formed by displacement of the terminal D-alanine using a serine hydroxyl group on the enzyme (Fig. 6). The acyl-D-alanyl residue was then transferred from this ester intermediate to the free amino group of an acceptor substrate. It was proposed that penicillin may be an active site-directed inhibitor capable of acylating bacterial transpeptidases because of the relatively high reactivity of its I)-lactam bond (Fig. 6). At that time the relationship of this primary event to cell death and lysis was predicted to be a direct consequence of weakening of the peptidoglycan. These effects are now known to be indirect because disruption of cell-wall biosynthesis is thought to activate “autolysins” which are peptidoglycan hydrolases that hydrolyse the cell wall causing lysis (Tomasz, 1979). Following Tipper’s and Strominger’s suggestions, there followed an intensive period of activity to purify and identify the targets for penicillin action. Exposure of bacterial plasma membrane preparations to radioactively labelled penicillin gives protein complexes covalently linked to the antibiotic. All bacterial membranes give rise to these complexes which can be detected by sodium dodecyl sulphate (SDS) -gel electrophoretic separation and

(h)

I ( & Dy-

RCONH

RCONH

ENZ-OH

0

0

C0;

OENZ

RCONH

7 ol-(N< 0-

H

C0;

'

co'

Inhibition product?

Figure 6

FIG. 6 The transpeptidation and hydrolysis reactions of N-acylated D-alanyl-D-alamine catalysed by serine enzymes ( a ) compared with a p-lactam antibiotic's reaction with serine enzymes ( b )

180

MICHAEL I PAGE

detection of the radioactively labelled proteins by autoradiography. This procedure was developed by Spratt (1975) and subsequently resulted in the detection of a number of bacterial proteins capable of binding penicillin which vary in number, molecular size and binding ability from one bacterium to another. For example, seven penicillin binding proteins (PBPs) have been isolated from E. coli (Spratt, 1977) as shown in Table 2 (Waxman and Strominger. 1983). The numerical connotation of PBPs refers to their relative molecular size within the group of PBPs detected in a microorganism (PBP 1 being the slowest moving on the gel and having the highest molecular size). PBP 1 of E. coli is not necessarily similar to PBP 1 of gonococci. PBPs constitute only 1 "h of the membrane-bound protein and number lo3 to lo4 per cell. They vary in type from three to ten per bacterium and in molecular weight from 30 000 to 100 000 (Spratt, 1980). PBPs fall into two main categories: the first comprises abundant, relatively low molecular weight proteins with in v i m D,D-carboxypeptidase activity, and the second, much less abundant, high molecular weight components which are difficult to purify and which frequently lack demonstrable in vitro activities. P-Lactam antibiotics appear to be lethal to growing E. coli cells by binding to and inactivating the "penicillin-sensitive enzymes" which correspond to the high molecular weight proteins PBP IA, IV, 2 and 3. All of these PBPs appear to be transpeptidases in vitro and all, but PBP 2, have also been shown to be transglycosylases in virro (Tipper, 1985). The situation in Gram-positive bacteria is more complex because of the wide variation in peptidoglycan structure (Ghuysen, 1977). The major PBP is a low molecular weight D,D-carboxypeptidase but in none has it been demonstrated unequivocally to be the lethal target. The selective inactivation of individual PBPs leads to strikingly unique morphological changes (Table 2) which could imply that these proteins perform a range of distinct physiological functions. The distinction between penicillin-sensitive enzymes (PSEs) and PBPs appears arbitrary and is probably a reflection of the interests of the investigators. Three non-membrane bound water-soluble D-alanyl-D-alanine peptidases have been studied in detail (Frkre and Joris, 1985): R39, R61 and alhus G enzymes. The R61 and alhus G enzymes have been crystallised and the crystal structure for R61 solved to 2.8 8, resolution (Kelly et al., 1985). Although for both R61 and R39 it is very unlikely that these exocellular enzymes are the killing targets of p-lactam antibiotics (Dusart et af., 1973), their study has provided useful models for the mechanism of inhibition of the enzymes anchored in the bacterial plasma membrane. The albus G enzyme contains zinc(I1) and catalyses the hydrolysis of the C-terminal D-alanine of D-alanyl-D-alanine terminated peptides but it is only very weakly inactivated by penicillins and cephalosporins (Frere and Joris, 1985).

TABLE2 Penicillin Binding Proteins (PBPs) of E. coli: properties and roles in cell-wall metabolism"

Apparent M.W. PBP (daltons)

Abundance Morphological effects ("/o total of inactivation PBPs) (where known)

Activities (where known)

Cell-survival/ viability

Proposed in vivo function

Non-essential

Minor transpeptidase, can compensate for PBP 1Bs

1A

90000

6

-

-

IB

87000

2

Rapid lysis

Transpeptidase and Essential possible transgl ycosylase

2

66000

0.7

Ovoid cell formation

-

3

60000

2

Filamentation

4

49000

4

-

5

42000

65

6

40000

21

a

Waxman and Strominger, 1983.

Major transpeptidase of cell elongation

Essential

Cell shape determination

Essential

Implicated in cell division and specifically in cross-wall formation

Carboxypeptidase Transpeptidase Endopeptidase

Non-essential

Secondary transpeptidase to increase cross-linking

-

Carboxypeptidase (transpeptidase)

Non-essential

Regulation of crosslinking

-

Carboxypeptidase (transpeptidase)

Non-essential

182

MICHAEL I. PAGE

The overall mechanism of interaction of the R39 and R61 enzymes (E) with p-lactams (L) can be represented by (1). The rate constant for the E

+L

K

EL&

k

EI

-+ k.3

E

products

(1)

irreversible formation of the enzyme-inhibitor complex (EI) is k, and the association constant for the reversible binding of the p-lactam to the enzyme is K . The enzyme may be slowly regenerated with a rate constant k,. To be an efficient inactivator of the transpeptidase enzymes the p-lactam should show a high value of k,K and a low value of k,. There have been several reports on the values of these kinetic constants with different p-lactam antibiotics (Frkre et af., 1975a; 1982, 1984; Kelly et al., 1981; Laurent et af., 1984). Although caution has been advised when examining the individual values of K and k, (Jencks and Page, 1972; Jencks 1975; Page, I980a, I98 I , 1984b), it has been tempting to interpret K as a measure of the strength of binding between the enzyme and substrate. Consequently the observed low association constants have been taken as evidence that recognition of the P-lactam inhibitor by the enzyme is very inefficient. Similar and low values of K d o not necessarily imply that discrimination between a group of substrates is poor. It is misleading to separate the values of K and k, because some of the intrinsic binding energy between the enzyme and substrate may be used to compensate for unfavourable energy changes which are necessary for efficient reaction within the enzyme-substrate complex (Jencks 1975; Page 1984b). Hence, a high intrinsic binding energy between enzyme and substrate can actually appear not as a high observed or apparent association constant but as an increased rate of decomposition of the enzyme-substrate complex. This could explain the variation in k, of up to lo5 with p-lactam structure. Furthermore, in vivo specificity results from a competition between substrates/inhibitors for the active site of the enzyme. The important parameter then becomes the free energy of activation as measured by k,K. Specificity between competing substrates is given by the relative values of k,K and not by the individual values of k, or K . In fact k,K varies by up to lo8 over a range of P-lactams with the R39 and R61 transpeptidase enzymes (Frere and Joris, 1985). This variation emphasises one of the difficulties of finding relationships between structure, chemical reactivity and activity towards enzymes. The R61 and R39 enzymes have been shown to have active site serine residues which are acylated by penicillin and other p-lactams (Georgopapadakou el a/., 1981), to give penicilloyl enzymes [13] as shown in Scheme 2 (Frere et at., 1976b; Degelaen et al., 1979b). Interestingly, ester [13] does not hydrolyse to give penicilloic acid [14], although this does occur if the

Vl

z

d

z 0 o

0

I r4

I

CI

+

3:

-

/

o=o2 z

0

d

/’’ 0

I 0

U

L

Y

0

184

MICHAEL I. PAGE

enzyme-complex is first denatured. At neutral pH the ester intermediate undergoes an enzyme catalysed cleavage of the C(5)-C(6) bond and the identified products are an acylated glycine [ 151 and N-formylpenicillamine [I61 (Frere et al., 1974, 1975b, 1976a, 1978; Adriaens et at., 1978). The precursor of the isolated [I61 is not known although a thiazoline is a possibility. All R61 and R39 enzyme-P-lactam complexes d o not degrade by this C(5)-C(6) fragmentation pathway. The reaction of the R61 enzyme with cephalosporins appears to give simple hydrolysis products (Kelly et at., 1981). In summary, the main mechanistic studies of p-lactam antibiotics with penicillin-sensitive enzymes have been performed with soluble exocellular enzymes which act as hydrolytic catalysts. Unfortunately, mechanistic studies with membrane-bound PBPs which act as penicillin-sensitive enzymes are far less advanced and so the molecular mechanism for the lethal action of p-lactam antibiotics remains speculative. 3 Is the antibiotic p-lactam unusual?

Because of the relatively rare occurrence of p-lactams in nature, it is not surprising that the biological activity of these compounds should be attributed to the chemical reactivity of the p-lactam ring. Shortly after the introduction of penicillin to the medical world it was suggested that the antibiotic's activity was due to the inherent strain of the four-membered ring (Strominger, 1967) or to reduced amide resonance (Woodward, 1949). The latter is conceivable because the butterfly shape of the penicillin molecule [I71 prevents the normal planar arrangement of the 0, C and N atoms assumed to be necessary for the effective delocalisation of the nitrogen lone pair. Both of these ideas are, of course, intuitively appealing and they have remained unchallenged for several decades. Indeed, these two proposals have dominated the thoughts of synthetic chemists who were, and to some extent still are, convinced that more effective antibiotics may be made by making the p-lactam system more strained or non-planar. However, the evidence to support an unusually strained or an amide resonance inhibited p-lactam in penicillin is ambiguous. The assessment of the nature of the p-lactam ring in penicillins and in cephalosporins and of its contribution to the reactivity of these molecules has been based on structural and kinetic studies. Before reviewing these, it should be noted that the potential contributions of strain energy release and inhibition of amide resonance are not trivial. I t is estimated that resonance stabilises amides by about 18 kcalmol-' (Fersht and Requena, 1971). The reason for the greater stability of amides compared with other carbonyl containing functionalities is attributed to the


185

RCO,

NH

unfavourable loss of resonance when nucleophiles attack the amide carbonyl carbon to form a tetrahedral intermediate (Scheme 3). If delocalisation is completely inhibited in an amide then the rate of a reaction, which would normally involve loss of resonance stabilisation in the transition state, could occur up to I0l3-fold (antilog 18/2.303 RT) faster than the analogous resonance stabilised system. The strain energy of a four membered ring is 26-29 kcal mol-’ (Page, 1973) and therefore a reaction involving opening of the p-lactam ring could take place faster by a factor of up to lozo (antilog 27.5/2.303 RT) than the analogous bond fission process in a strain-free amide. If strain or resonance inhibition is even slightly significant in penicillins and cephalosporins their effects should therefore be easily observable.

..

Nu

+

\C-{j(

--.+

6’

-0-k-N:

I + Nu Scheme 3

The treatment of amide resonance as a result of delocalisation of the nitrogen lone pair by overlap of the 2p-orbitals on the participating atoms [I81 has been used to predict that a pyramidal amide nitrogen will cause loss of resonance energy (Woodward, 1949). Pyramidalisation of the nitrogen will, however, not necessarily produce the same effect as that caused by rotation about the C-N bond, i.e. a change in the dihedral angle between the p-orbitals with the 0, C and N atoms remaining coplanar. These contrasting effects are illustrated, respectively, by the Newman projection formulae [19] and [20] obtained by looking along the C-N bond. Despite these differences, the assumption that amide resonance in penicillins and cephalosporins is inhibited has been generally accepted and several experimental observations have been used to support this suggestion.

186

MICHAEL I PAGE

S T R U C T U R A L A N D G R O U N D - S T A T E EFFECTS

Amide resonance is usually depicted by the canonical forms [21] and [22].

Inhibition of amide resonance should make the amide resemble [21] at the expense of [22]. The reasonable conclusion would be that, compared with a normal amide, resonance inhibition would: (i) increase the C-N bond length and decrease the C-N bond strength. (ii) decrease the C-0 bond length and increase the C-0 bond strength. (iii) decrease the positive charge density on nitrogen. (iv) decrease the negative charge density on oxygen. Planarity of the nitrogen and bond lengths X-ray crystallography has been invaluable in providing detailed three-dimensional structures of the p-lactam antibiotics. Despite the obvious criticism that solid state conformations may not be relevant to solution conformations, the geometrical picture presented from X-ray data has been the basis for many discussions relating chemical structure and biological activity. Although acetamide has C , symmetry in the gas phase and solution, the carbonyl carbon is pyramidalised in the crystalline state (Jeffrey et al., 1980). The degree of coplanarity of the p-lactam nitrogen with its three substituents can be expressed either by the perpendicular distance, h, of the nitrogen from the plane of its substituents or by the sum of the bond angles about nitrogen. The former is easier to visualise and the nitrogen ranges from being essentially in the plane of its three substituents in monocyclic p-lactams to being 0.5 8, out of the plane in bicyclic systems. Examples of /I-values are given in Table 3 and there have been several reviews on


187

TABLE3 Structural parameters of some p-lactams

Compound

G O stretch/ cm-'

Pmicillitu" ampicillin' benzy lpenicillin' phenoxymethylpenicillind

1770-1790

A3-Cephalosporinse cephaloridinel

1760-1 790

A'- Cephcilosporinse phenoxymethyl A'cephalosporinf

1750-1 780

Anhydropmici/hfl phenoxymethylanhydropenicillinh

1810

Monocyclic p-lactams

1730- 1760

Aniides

1600-1680

Morris and Jackson, 1970 James et a/., 1968 ' Pitt, 1952 Abrahamson et al.. 1963

Distance of N from lane h J

p-lactam p-lactam C=O bond C-N bond length/A length/A

0.38 0.40 0.40

I .20 .I7 .21

I .36 1.34 1.46

0.24

.2 1

1.38

0.06

.22

I .34

0.41

1.18

1.42

0 0

1.21 1.24

1.35 1.33

'' Green ei d., 1965 Sweet and Dahl, 1970 Wolfe e t d . , 1968 Simon et al.. 1972

structural data determined by X-rays (Sweet and Dahl, 1970; Simon et a/., 1972; Sweet, 1973; Boyd, 1982). Until recently it had been generally assumed that a more pyramidal nitrogen decreased amide resonance and increased biological activity. Great effort was therefore discharged in making non-planar p-lactams. A I-carba-1-penem shows the highest h-value, 0.54 8, (Woodward, 1980), and yet this compound is biologically inactive. Furthermore, there is no direct correlation between h-values and chemical reactivity (see Section 4). It has been claimed that bond length data for penicillins and cephalosporins show evidence for the inhibition of amide resonance (Sweet, 1973). The C-N bond length of planar monocyclic p-lactams (1.35 A) is generally longer than that of amides (1.33 A). The converse is true f o r C-0 bond lengths, 1.24A for amides compared with 1.21 8, for monocyclic p-lactams. In non-planar penicillins and cephalosporins there is a general trend for the C-N bond length to increase as the C=O bond length decreases (Sweet, 1973; Simon et al., 1972). However, this trend is by no means linear. Bond lengths for C-0 vary from 1.17 to 1.24 A and for C-N

188

MICHAEL I. PAGE

from 1.33 to I .46 8,. There is also a tendency for the C-N bond length to increase with h. It is difficult to discern reasons and consequences of these bond length differences. Penicillin V [ I ; R = PhOCH,] shows the longest C-N bond length of 1.46 8, (Abrahamsson et a / . , 1963) and yet the C==O bond length is identical to that commonly found in planar monocyclic p-lactams ( I .2 I 8,). In monocyclic p-lactams the nitrogen is coplanar with its three substituents and yet the bond length differences are also in the direction predicted by inhibition of amide resonance. The degrees of non-planarity in penicillin V [ I ; R = PhOCH,] and ampicillin [ I ; R = PhCH(NH,)] are similar ( h = 0.40 and 0.388,, respectively) and yet the C-N bond length in the former is 0.108, longer than in the latter (Abrahamsson et a / . , 1963; James et a/., 1968).

[23]

Structural data have also been used to support the suggestion that enamine resonance [23] is important in cephalosporins and that this also reduces amide resonance (Sweet, 1973). However, there is no significant difference in the C-0 and C-N bond lengths of cephalosporins from that general trend exhibited by penicillins. Although it is claimed (Sweet, 1973) that the C(4)-N(5) of A3-cephalosporins is significantly shorter than the “expected” values this is not generally true. For example, C(4)-N(5) of cephaloglycine is 1.51 8, (Sweet and Dahl, 1970) which is longer than the quoted expected value of 1.47 8, for C-N. Furthermore, the C(4)-N(5) bond length in the A3-cephalosporin, cephaloglycin, is longer than that of 1.45 8, in A2-cephems (Sweet and Dahl, 1970) and that of 1.46 8, in cephams (Vijayan e f a / . , 1973) where enamine resonance cannot occur. It would seem logical to conclude that variations in bond lengths within penicillin and cephalosporin derivatives are caused by the nature of substituents and the minimisation of unfavourable strain energies caused by the geometry of the molecule. To attribute these differences to the inhibition of amide resonance seems speculative and is only supported by the selection of examples. Nmr chemical shifts The conformation of the substituent on nitrogen relative to the carbonyl group has a significant effect on the carbonyl 13C chemical shift in amides. For example, a difference of 4ppm is observed in the carbonyl 13C


189

resonances of the E- and Z-isomers of N-methylformamide (Levy and Nelson, 1972a). This is a reflection of both steric and anisotropic differences in the environment. There does not appear to be a simple relationship between 3C chemical shifts and the local electron density distribution. However, there have been several attempts to correlate antibacterial activity with chemical shifts. The p-lactam carbonyl carbon usually resonates between 16G167 ppm in a 13C nmr spectrum (Bose and Srinivasan, 1979). This is in the same region in which the carbonyl resonances of formamide and its N-methylated derivatives also appear (Levy and Nelson, 1972b). It is interesting to note that the carbonyl resonances of 5- or larger-membered lactams appear between 170 and 180 ppm (Williamson and Roberts, 1976). Replacement of an alkyl substituent on the 0-lactam nitrogen by an aryl substituent causes shielding of the lactam carbonyl resonance by ca 4 ppm (Bose and Srinivasan, 1979). Dipole moment (Malihowski et al., 1974) and uv spectral studies (Manhas et al., 1968) indicate that the lone pair on the p-lactam nitrogen is conjugated with the aromatic ring [24]. This resonance interaction presumably alters the electron density at the p-lactam carbonyl and could account for the shielding of the resonance due to that carbon.

There is little variation in the chemical shifts of the p-lactam carbonyl carbon of penicillins and cephalosporins (Mondelli and Ventura, 1977; Paschal et al., 1978; Dereppe et a1 1978; Schanck et al., 1979). The carbonyl carbon of the p-lactam in penicillins resonates about 10 ppm to lower field than that in cephalosporins. Surprisingly, the shifts in the biologically active A3- and the inactive A’- cephalosporins are similar (Mondelli and Ventura, 1977). Inhibition of amide resonance may be expected to make the carbonyl carbon more electron deficient (Dhami and Stothers, 1964). Although the difference in chemical shifts between penicillins and A3-cephalosporins support this proposal if amide resonance is inhibited by a pyramidal p-lactam nitrogen, it is not apparent from the A’-/A3- cephalosporin comparison. In penicillins the nitrogen is 0.4A from the plane of its substituents compared with 0.2A in A3-cephalosporins, whereas the ceph-2-em systems are planar (Flynn, 1972). The similarity of the values of the I3C shifts found for the (j-lactam carbonyl carbons in ceph-3-ems and

190

MICHAEL I. PAGE

ceph-2-ems indicates that the charge density and bond order at the carbonyl carbons in both systems is approximately the same. This argues against enamine type resonance [23] in ceph-3-ems. "N-Chemical shifts of the p-lactam nitrogen in ceph-3-ems are almost 1 ppm) with the nature of the substituent at C(3) (Paschal et invariant ( < a/., 1978) and therefore also d o not indicate significant enamine type resonance in these systems. Not surprisingly there is a large difference of 15 ppm in the "N-chemical shifts of the p-lactam nitrogen in ceph-2-ems and ceph-3-ems. Interestingly, there is an upfield shift of 30ppm in the p-lactam nitrogen on going from non-planar penicillins to planar ceph-2-ems (Lichter and Dorman, 1976). Increased amide conjugation in the planar system would be expected to have induced a downfield shift. No doubt, since the factors determining l5N-chernical shifts are not very well understood, it could be argued that the shifts are insensitive to minor changes in the degree of amide delocalisation. Infrared carbon-yl stretching frequency The p-lactam infrared stretching frequency ( v C = J has been regarded as an important index for both inhibition of amide resonance and for investigating structure-activity relationships of the p-lactam antibiotics (Morin et al., 1969; Sweet and Dahl. 1970; Demarco and Nagarajan, 1973; Indelicato et a/., 1974; Murakami ct al., 1980; Takasuka et al., 1982; Nishikawa et al., 1982). In normal penams the p-lactam carbonyl stretching frequency occurs in the 1770-1790cm-' range compared with 1730-1760cm-' for monocyclic unfused p-lactams and about 1600-I 680 cm- ' for amides. In general, the non-planar 3-cephems show higher stretching frequencies, 1 7 8 6 1790 cm- ', than the planar 2-cephems which absorb at 1750-1 780 cm-'. The frequency in cephalosporins increases by ca 5 cm- when the ring sulphur is replaced by oxygen but decreases by a similar amount when the 7-a-hydrogen is substituted by a methoxy group (Takasuka et a/., 1982). It is difficult to make generalisations about the observed p-lactam frequency since different conditions of measurement (KBr, film, solution, etc.) may cause variations comparable with those produced by structural changes. There is a tendency for a high carbonyl stretching frequency to be associated with a shorter p-lactam C=O bond length and a more pyramidal nitrogen. Furthermore it has been tempting to associate a high carbonyl stretching frequency with increased strain, increased double bond character and reduced amide resonance (Morin et d., 1969; Simon et d., 1972; Pfaendler et d., 1981). However, the evidence again is ambiguous. Although selected examples may show some of these interrelationships there are many exceptions; for example, the carbonyl stretching frequency for some penems decreases

'

MECHANISMS OF REACTIONS OF p - L A C T A M ANTIBIOTICS

191

20 cm- whilst the 0-lactam nitrogen becomes more pyramidal by 0.12 A (Pfaendler et al., 1981). The direct interpretation of carbonyl stretching frequencies in terms of bond strengths or electron density distributions is not straightforward. Many subtle effects can alter the frequency even if the force constant for G O stretching, which is presumably the best indicator of bond strength, remains constant. For example, in the system X - W the carbonyl stretching frequency can be increased by decreasing the C-X bond length, by increasing the C-X stretching force constant or by increasing the XCO bending force constant (Collings et al., 1970). Ultraviolet absorption and circular dichroism spectra The 3-cephem systems show two characteristic absorptions in their uv absorption spectra, a strong band near 260nm and a weaker one at about 230 nm. The chromophore near 260 nm is not associated with the presence of sulphur (Wolfe et al., 1974) and it appears that the amide and enamine systems contribute to this absorption. Semi-empirical MO calculations (Boyd, 1972) suggest that the 260nm absorption is associated with the excitation from the ground state with a doubly occupied enamine K MO to an excited state described mainly by a configuration with the excited electron in an MO with both K*(C==O)and n*(C=C) character. Circular dichroism (CD) spectra of penams show a strong positive Cotton effect at about 230 nm and a weaker one of variable sign at about 207 nm (Busson et al., 1976a; Richardson et al., 1977). A C D study of a I-carbapenam, (559- 1-azabicyclo[3.2.0]heptan-7-one, in which the p-lactam chromophore is free of sulphur, carboxyl and other amido groups, also shows the Cotton effect at 231 nm (Busson and Vanderhaeghe, 1978). It is unlikely therefore that the band in penicillins is due to charge transfer from sulphur to amide as suggested by MO calculations (Boyd et al., 1976). Furthermore, monocyclic 0-lactams show similar bands (Busson et al., 1978) and the simplest interpretation of the observations is that the Cotton effects are associated with n + K* transitions characteristic of all amides. Theoretical calculations Theoretical geometry optimisation of 0-lactams at a semi-empirical level and a limited ah initio study using minimal basis STO-3G calculations at fixed geometries have been reported (Glidewell and Mollison, 198I). The STO-3G basis set underestimates valence angles at heteroatoms and is expected to exaggerate the degree of non-planarity in amides whilst the split valence 4-3 1G basis set characteristically overestimates these valence angles and consequently overestimates the tendency of amides to be planar. Given these reservations the calculated STO-3G energy of formamide in a

192

MICHAEL I. PAGE

penicillin-like geometry is only 2.8 kcal mol-' higher than the planar geometry (Vishveshwara and Rao, 1983). Furthermore, in general, the geometrical parameters associated with the p-lactam ring vary only slightly with changes in the hybridisation at nitrogen. An exception is the C-N bond length which becomes longer as the nitrogen becomes pyramidal. The barrier to inversion at nitrogen in ammonia is 5.8 kcal mol-' which is much greater than that in molecules like formamide (Radom and Riggs, 1980; Carlsen et al., 1979). Formamide lies in a potential well which is very flat with respect to inversion at nitrogen. The inversion barrier is lower for molecules favouring a large angle at nitrogen (amides) and higher for systems adopting a small angle at nitrogen (e.g. aziridine) (Stackhouse et al., 1971). It appears that the nitrogen in amides can be made pyramidal without severe changes in energy. Calculations of charge distributions of p-lactam rings, and in particular the Mulliken overlap populations for the CO and C N bonds, have been reported for several systems (Boyd, 1982). The calculated charge density on nitrogen becomes less negative, by 0.013 of atomic charge, as the nitrogen of a p-lactam is made more pyramidal. However, the calculated charge on the carbonyl carbon becomes less positive as the nitrogen becomes pyramidal (Glidewell and Mollison, 1981) which would presumably suggest that p-lactams with a pyramidal nitrogen are less susceptible to nucleophilic at tack. Basicity of p-lactam nitrogen Inhibition of amide resonance in bicyclic p-lactams will make the amide resemble canonical form [21] at the expense of [22]. A consequence of increased localisation of the lone pair on nitrogen would be to increase the basicity of nitrogen. It is well known that torsional strain in amides can increase the basicity of nitrogen in amides. For example, 6,6-dimethyl- 1-azabicyclo[2.2.2]octan2-one [25] presumably has the nitrogen lone pair almost orthogonal to the carbonyl IT system and amide resonance is consequently inhibited. The carbonyl absorption for [25] is at 1762 cm-' (Pracejus et al., 1965) which lies

PSI

between the normal amide carbonyl stretch and that for some N-acetyltrialkylammonium tetrafluoroborates at about 1816 cm- (Paukstelis and Kim, 1974). Amides are normally only very weakly basic and the pK,-values

'


193

of their conjugate acids are around zero. By contrast, [25] is half protonated at pH 5.3, consistent with the increased basicity of the amide nitrogen. If amide resonance in penicillins is inhibited because of the pyramidal nature of the p-lactam nitrogen, penicillins should also show enhanced basicity compared with normal amides. There is no evidence to suggest that this is the case. In fact, penicillins appear to show reduced basicity and cannot be detectably protonated even in 1 2 M hydrochloric acid. For N-protonation of penicillins pK, must be < -5 (Proctor et al., 1982). This reduced basicity is discussed in Section 5. Another indication of increased nitrogen basicity would be a large binding constant of penicillin to metal ions. However, the equilibrium constant for metal-ion coordination between the carboxyl group and p-lactam nitrogen [26] is only about 100-200 M - ' for copper(II), zinc(II), nickel(I1) and cobalt(I1) (Gensmantel et al., 1978, 1980). This is the order of magnitude expected for coordination between a normal amide and a carboxyl group.

[W There appears, therefore, to be no evidence of substantial inhibited amide resonance in penicillins and the p-lactam nitrogen shows no enhanced electron pair donating ability to either a proton or to metal ions. Resonance and non-planarity The degree of amide resonance that is lost by distortion of the normal planar geometry is not immediately apparent. The angular dependence of resonance interactions, particularly steric inhibition of resonance, has been of long-term interest (Hammond, 1956; Wheland, 1955; Wittig and Steinhoff, 1964; Dewhirst and Cram, 1958; Ree and Martin, 1970; Buss el al., 1971; Doering et al., 1984). The importance of residual resonance stabilisation even when the interacting p-orbitals are orthogonal is highlighted by ab initio studies. For example, nearly 50% of the possible charge donation from oxygen to the ring carbons in phenol occurs even in the perpendicular conformation (Hehre et al., 1972). Rotation about bonds which increases the dihedral angle between porbitals on adjacent atoms does decrease resonance interaction. For example, the increased basicity of the quinuclidin-2-one [25] may be attributed to decreased amide resonance. Similarly, the energy barrier to rotation about

194

MICHAEL I. PAGE

the C-N bond in amides of 2G30 kcalmol-’ (Kessler, 1970; Stewart and Siddall, 1970) indicates a significant torsional dependence for resonance. However, resonance interaction appears to remain significant if nitrogen adjacent to a 71 or p system is made pyramidal. For example, 7c-sp3 conjugation in aromatic amines can be appreciable and the configuration about nitrogen in most simple anilines is nearly pyramidal rather than trigonal (Bottini and Nash, 1962; Aroney et al., 1968). The geometry of the bicyclo[3.3.Ilnonane system [27] does not completely inhibit resonance stabilisation of the developing bridgehead cation in solvolysis reactions (Krabbenhoft et al., 1974). The a-amino substitutent in [27; Y = NCH,] increases reactivity relative to the carbon analogue [27; Y = CH,] by a factor of 10’. The geometry of the 1-adamantyl-type cation [28] imposes a perpendicular twist on the 2-heteroatom’s lone pair relative to the vacant carbon p-orbital and would be expected to minimise resonance stabilisation.

However, despite this unfavourable geometry there is still significant interaction ( I 8% of 0; used to correlate rates) between the cation and non-bonding electrons of the nitrogen, and [28; Y = NCH,] is 356-fold more reactive than [28; Y = CH,] (Meyer and Martin, 1976). Only in the case of the much more rigid bicyclo[2.2. Ilheptane system [29] has an a-amino substituent been found to decrease reactivity (20-fold) (Gassman et al., 1972) but even here resonance stabilisation must still occur because an a-nitrogen is expected inductively to destabilise a carbocation leading to a decrease in reactivity by a factor of ca lo5. So even though a carbocation, because of its lower stability, probably makes a greater and less stereospecific demand upon adjacent nitrogen lone pairs than a carbonyl group these observations suggest that pyramidal distortion of nitrogen in amides will not cause significant loss of resonance. Even though the nitrogen in penicillins is pyramidal there is little evidence from ground-state effects to support the often quoted claim that this causes significant inhibition of amide resonance. The next section shows that kinetic observations are in agreement with this conclusion. KINETIC EFFECTS

Nucleophilic substitution at the carbonyl group of an amide invariably occurs in a stepwise manner by initial formation of a tetrahedral interme-

195

MECHANISMS OF REACTIONS OF a-LACTAM ANTIBIOTICS

diate (Scheme 3, p. 185). Conversion of the three-coordinate, sp2-hybridised carbonyl carbon to a four-coordinate sp3-hybridised carbon in the intermediate must be accompanied by the loss of amide resonance. This contribution to the activation energy will be reduced if amide resonance is inhibited and a rate enhancement is expected in such cases. Similarly, the release of strain energy will increase the rate if the four-membered ring is opened or has been opened in the transition state. The total strain energy of four-membered rings is probably not released until there is significant bond extension (Earl et ul., 1983). A simple reaction to see if either of these effects is apparent is the hydrolysis of the P-lactam antibiotics. The alkaline hydrolysis of benzylpenicillin opens the p-lactam ring to give benzylpenicilloate [30]. With respect to hydroxide ion, benzylpenicillin shows a reactivity similar to that of ethyl acetate. The pKa-value of the protonated amine in the thiazolidine derivative [30] is 5.2 and because of this weakly basic nitrogen the leaving group ability of the amine is expected to be improved. Therefore, in order to assess any special reactivity of the p-lactam antibiotics, the dependence of the rate of hydrolysis of simple amides and P-lactams upon substituents must be known.

r0

/

\

/-I/--

0-C-NHR

A Bransted plot of the second order rate constants for the hydroxide ion catalysed hydrolysis of acyclic amides, monocyclic P-lactams and bicyclic P-lactams is shown in Fig. 7 (Proctor et al., 1982). The Bransted PI, value for N-substituted acyclic amides and anilides is -0.07. Since the nitrogen of amides has an effective charge of +0.6 in the ground state, this is indicative of a transition state in which the nitrogen behaves as if it has ca 0.5 positive charge in the transition state (Morris and Page, 1980~).The alkaline hydrolysis of acyclic anilides is characterised by "0 exchange with solvent occurring faster than hydrolysis (Bender and Thomas, 1961) and by a second-order dependence on hydroxide ion concentration (Schowen et al., 1966). Both results indicate rate-limiting breakdown of the tetrahedral intermediate. The Bransted PI, is incompatible with rate-limiting expulsion of the amine anion or with a transition state in which the nitrogen has a unit positive charge, i.e. is fully protonated. The observations are consistent with water acting as a general acid catalyst in the breakdown of the tetrahedral intermediate [31] (Morris and Page, 1980~;Proctor et al., 1982). The rates of alkaline hydrolysis of p-lactams exhibit a first-order dependence on hydroxide ion concentration and show a Bransted PI, value of -0.44, which is

196

MICHAEL I. PAGE

M-ls-l lo%! H

1.0

0.0

- 1.0 -2.0

- 3.0 - 4.0 - 5.0

- 6.0 1

1

2

3

It

5 6 I 8 p K , o f leaving group

9

10

11

12

FIG.7 Brnnsted plot of the second-order rate constants for the hydroxide-ion catalysed hydrolysis of acyclic amides ( ), monocyclic p-lactams (0),and bicyclic p-lactams ( A ) against p K , for the leaving group amine. Data refer to 30°C and are taken from Proctor er al., 1982

+

indicative of rate-limiting formation of the tetrahedral intermediate (Blackburn and Plackett, 1972). However, p-lactams of the more basic amines show a positive deviation from this line and exhibit a smaller dependence upon basicity of the leaving group which may signify a change in rate-limiting step to breakdown of the intermediate. (Proctor et af., 1982). A consequence of the different dependence upon leaving group basicity is that the rate enhancement of p-lactams compared with acyclic amides depends upon the basicity of the leaving group amine. P-Lactams of weakly basic amines are ca 500-fold more reactive than an acyclic amide of the same


197

amine. However, p-lactams of basic amines are only slightly more reactive than an analogous acyclic amide. Similarly, p-lactones are only about 10-fold more reactive than analogous esters (Blackburn and Dodds, 1974). Crystallographic (Luche et a[., 1968; Parthasarathy, 1970; Chambers and Doedens, 1980) and spectroscopic evidence (Manhas et al., 1968) show that N-substituted p-lactams are planar and resonance-stabilised as in amides. The rate enhancement of 30-500-fold shown by p-lactams of amines of pKa < 6 may be adequately rationalised by the change in coordination number/hybridisation of the carbonyl carbon as the tetrahedral intermediate is formed (Page, 1973; Gensmantel et al., 1981). The magnitude is similar to the 500-fold faster rate of reduction of cyclobutanone by borohydride compared with acetone (Brown and Ichikawa, 1957). The conversion of three-coordinate to four-coordinate carbon in four-membered rings is accompanied by the release of 11.4 kJmol-' of strain energy (Page, 1973; Allinger et al., 1972). A rate enhancement of 100-fold can therefore be expected in the conversion of the p-lactam carbonyl carbon to a tetrahedral intermediate. The change in strain energy will be even greater for a resonance stabilised p-lactam which has some endocyclic double bond character within the four-membered ring. As the rate-limiting steps for the alkaline hydrolysis of amides and p-lactams are different, the relatively small rate enhancement shown by p-lactams indicates that the energy of the transition state for breakdown of the tetrahedral intermediate in amide hydrolysis is not significantly greater than that for formation of the intermediate. There is nothing unusual about the chemical reactivity of the monocyclic p-lactam antibiotics nocardicin [8] and the monobactams [9]. The second order rate constants for their alkaline hydrolysis fit the Brsnsted plot (Fig. 7) (Proctor et al., 1982). 0-Lactams of basic amines (pKa of RfiH, > 7) are only ca 10-fold more reactive than analogous acyclic amides. The Brsnsted p,, values for both of these systems are compatible with rate-limiting breakdown of the tetrahedral intermediate. It follows that there can be little or no 0-lactam C-N bond fission in the transition state. The release of strain energy accompanying the opening of the p-lactam ring could increase the rate by up to lozo SO this energy must still be present in the transition state for the hydrolysis of p-lactams. Fusing the p-lactam ring to a five-membered ring to make 1-aza-bicyclo[3.2.0]heptan-2-ones increases the reactivity by ca 100-fold but does not significantly change the Brsnsted p,, value which is -0.55 for the bicyclic system (Proctor et al., 1982). Although the rate enhancement is substantial, it is hardly of the magnitude expected from the release of strain energy in opening a four-membered ring or from a system in which amide

198

MICHAEL I. PAGE

resonance is significantly inhibited. Ring opening does not lower the activation energy because the rate-limiting step for the alkaline hydrolysis of penicillins is formation of the tetrahedral intermediate. The B r ~ n s t e dPI, of -0.55 indicates that the nitrogen behaves as if it has no charge in the transition state and has lost all of the expected 0.6 positive charge present in the resonance stabilised p-lactam. This is compatible with a transition state that very much resembles the tetrahedral intermediate. It cannot be argued that the transition state is ‘early’ and involves little bond formation between the carbonyl carbon and the attacking hydroxide ion and that therefore any effect of amide resonance inhibition is not manifest in the transition state. Furthermore, the PI, of -0.55 indicates that the positive charge density on the P-lactam nitrogen in penicillins is similar to that in monocyclic p-lactams and amides where resonance is established. It is worth noting that monocyclic p-lactams of weakly basic amines can be as chemically reactive as penicillins and cephalosporins. It is not necessary to make the p-lactam part of a bicyclic system to have a reactive amide. Although it has been suggested that the reactivity of penicillins may be due to intramolecular nucleophilic attack of the 6-acylamido side chain on the p-lactam carbonyl (Doyle and Nayler, 1964; Moll, 1968), there is no significant dependence of k o H - on the nature of the side chain (Yamana et d., I974b; Bundgaard, 1972). SUMMARY

Both kinetic and ground-state effects d o not indicate a significant degree of inhibition of amide resonance in penicillins and cephalosporins. The bicyclic p-lactam antibiotics d o not exhibit exceptional chemical reactivity. Monocyclic p-lactams with suitable electron withdrawing substituents may be as reactive as the bicyclic systems. A pyramidal geometry of the p-lactam nitrogen does not necessarily give a chemically more reactive p-lactam. Strained p-lactams are not necessarily better antibiotics and biological activity is not directly related to chemical reactivity. 4

Alkaline hydrolysis and structure: chemical reactivity relationships

I t is well known that minor substituent changes in p-lactam antibiotics can have a dramatic effect upon antibacterial activity and susceptibility to p-lactamase catalysed hydrolysis. In order to identify the binding energy effects between the p-lactam and enzymes, the effect of substituents on chemical reactivity must be known. The effects of structural changes on the rates of alkaline hydrolysis of penicillin and cephalosporin derivatives are summarised in Table 4.


199

TABLE4 Second-order rate constants (M - Is- I ) for the hydroxide-ion catalysed hydrolysis of penicillins, cephalosporins and other p-lactam antibiotics at 30°C; I = 1 .O M (KCI)"

r%

0'

C0;

PhCH,CONH

'42N5r"i. D3 0

0

C0;

CO,

7.40 x 10-3

6.31

1.54 x l o - '

x

PhCH,CONH

D9--""

Yl

0

COY

I

co;

2.39

PhCH,CONH

phcH2c0N* O * N F C H 3

co, 2.90 x

10-2

0

CH,

co, 1.10 x 10

PhCH,CONH

1.78 x 10 " Sec also Table 5. p. 205

I

200

MICHAEL I. PAGE

PENICILLINS

The mechanism of the alkaline hydrolysis was reviewed in the previous section and it has already been mentioned that decreasing the basicity of the leaving group amine increases the rate of alkaline hydrolysis of penicillins. The Br~nstedPI, of -0.55 (Proctor el al., 1982) is indicative of rate-limiting formation of a tetrahedral intermediate [32]. Electron withdrawing substituents at C(6) also increase the rate of hydroxide ion hydrolysis and give a Hammett p,-value of +2.0 (Proctor et al., 1982). This is only slightly less than the value of 2.7 reported for acyclic amides (Bruylants and Kezdy, 1960). Substituents at C(6) affect the rate of nucleophilic substitution by their effect upon both the electrophilicity of the carbonyl carbon and the leaving group amine. Although the acylamido side chain at C(6) is important for biological activity and increases the rate of alkaline hydrolysis 20-fold relative to penicillanic acid, its effect on chemical reactivity is purely inductive. 6-a-Chloropenicillanic acid undergoes alkaline hydrolysis 2.7 times faster than the 6-P-acylamido derivative. RCONH

OH -/

It is conceivable that the carboxyl group at C(3) which is also very important for biological activity, could act as a general acid catalyst in the reaction of nucleophiles with penicillins [33]. However, there is no evidence that the carboxyl group facilitates hydrolysis and, as expected on the basis of a purely inductive effect, the esterification of this group increases the rate of reaction 10- to 100-fold (Proctor et al., 1982; Gensmantel et al., 1978). The replacement of the thiazolidine S by CH, to give a carbapenam increases the rate by a factor of 3, whereas substitution by 0 as in the oxapenams increases the rate ca 5-fold (Table 4). The effect of the replacement by oxygen is that expected on the basis of an inductive effect. The CSC bond angle is relatively small and the CS bond length relatively long compared with the carbon analogue. Replacement of S by CH, will decrease the leaving group ability of the p-lactam amine [NH,CH,CH,SMe has pK, 9.2 compared with 10.6 for fiH,(CH,),Me]. Presumably these effects must cancel so that there is little difference in reactivity between the penams and carbapenams.

MECHANISMS OF REACTIONS OF b - L A C T A M ANTIBIOTICS

201

The incorporation of a double bond into the thiazolidine ring of a penam to give the corresponding penem system [7] increases the rate of hydrolysis by ca 25-fold. This is the order of magnitude expected from the decrease in basicity of the leaving group amine brought about by the introduction of a conjugated amine in the tetrahedral intermediate (Proctor et al., 1982). Conversion of a A'-carbapenem to a A '-carbapenem similarly decreases the reactivity 25-fold (Pfaendler et al., 1981). Anhydropenicillins [34] caused some anxiety to early workers because their apparent chemical stability could not be reconciled with the assumed strain in the system (Wolfe et al., 1963, 1968). The p-lactam nitrogen is pyramidal, h = 0.41 8, (Simon et al., 1972), and the carbonyl stretching frequency is higher than that found in normal penicillins. The characterisation of anhydropenicillins as chemically stable compared with penicillins was based on the observation that they were recovered unchanged from refluxing solvents such as water and ethanol (Wolfe et al., 1968). In fact, it was later shown (Bundgaard and Angelo, 1974) that anhydropenicillins are ca 100-fold more reactive towards alkaline hydrolysis than are normal penicillins. This is as expected because of the effect of the electron withdrawing sulphoxide group on the leaving group amine. Bundgaard and Angelo (1974) suggested that nucleophilic attack was on the thiolactone of [34], but it has since been demonstrated that this occurs on the p-lactam carbonyl (Pratt et al., 1983). The rate of the alkaline hydrolysis of penicillins is not greatly affected by ionic strength. The second-order rate constant increases by about 30% up to I = 0.5 M (KCI) but is then independent of I up to I = 4.0 M (Morris and Page, 1978). The observed pseudo first-order rate coefficient for the hydrolysis of benzylpenicillin is first order in hydroxide ion up to 2 M sodium hydroxide. Above this concentration it begins to level off (Minhas and Page; 1982). This is probably attributable to ionisation of the benzylamido side chain. Presumably hydroxide ion attack on the penicillin with a 6-amido anion side chain is retarded. In support of this, the observed rate constant for the hydrolysis of phenoxymethylpenicillin shows a non-linear dependence upon hydroxide ion above 0.1 M sodium hydroxide (Minhas and Page, 1982; Pratt et al., 1983). The more electron-withdrawing phenoxymethyl group decreases the pKa-value of the amide side chain to 13.3. The observed first-order rate constants for the hydrolysis of 6-aminopenicillanic acid are, as expected, linear in hydroxide ion concentration. Solvent isotope effects on the alkaline hydrolysis of penicillins are consistent with rate-limiting formation of the tetrahedral intermediate; koH-/koD- = 0.65 at 30°C and 0.59 at 35°C (Gensmantel et al., 1978; Yamana et al.,1977).

MICHAEL I. PAGE

202

CEPHALOSPORINS

The major structural differences between cephalosporins [ 1 I] and penicillins [ 121 are that the five-membered thiazolidine ring of penicillins is replaced by a six-membered dihydrothiazine ring in cephalosporins and that the degree of pyramidalisation of the p-lactam nitrogen is generally smaller in cephalosporins. In addition, many of the cephalosporins have a leaving group, e.g. acetate, pyridine and thiol, at C(3’) and expulsion of these groups occurs during the hydrolysis of the S-lactam as shown in Scheme 4 (HamiltonMiller et al., 1970a; O’Callaghan et a[., 1972). There have been many RCONH

RCONH

N”;. B&:i2-L

0

Op $ C H co; 2

0;

Scheme 4

+

1371

suggestions (Bundgaard, 1975), apparently supported by theoretical calculations (Boyd et al., 1975; Boyd and Lunn, 1979), that a nucleophilic attack on the p-lactam carbonyl carbon is concerted with departure of the leaving group at C(3’) [35]. It is common to read that the presence of the leaving group at C(3’) enhances chemical reactivity (Wei et al., 1983). Furthermore, it has been proposed that biological activity is related to the leaving group ability of the C(3’) substituent (Boyd et a/., 1980). RCONH

ws\ coy

In general, the second-order rate constants for the hydroxide-ion catalysed hydrolysis of cephalosporins are similar to those of penicillins (Proctor ef al., 1982; Yamana and Tsuji, 1976). This similarity indicates that the non-planarity of the p-lactam nitrogen does not significantly affect amide resonance since the nitrogen is 0.4A out of the plane defined by its substituents in penicillins (Sweet and Dahl, 1970) whereas in the cephalosporins it deviates by 0.2-0.3A (Sweet, 1973). The kinetic similarity also indicates that having a leaving group at C(3’) does not significantly affect the reactivity of cephalosporins. The rate-limiting step in the alkaline hydrolysis


203

of cephalosporins appears to be formation of the tetrahedral intermediate [32]. Electron-withdrawing substituents attached to the p-lactam nitrogen increase the rate of hydrolysis and give a B r ~ n s t e dPI, of -0.6 (Proctor et a/., 1982). In the stepwise mechanism, breakdown of the tetrahedral intermediate generates the enamine [36] followed by expulsion of the leaving group at C(3’) to give the conjugated imine [37] (Agathocleous et al., 1985). It has been observed that the rate of appearance of the leaving group L is identical to the rate of p-lactam ring opening (Coene et a/., 1984; Bundgaard. 1975).

COS [361

Although this has been interpreted as supporting the concerted mechanism, it is, of course, also consistent with a stepwise process involving rate-limiting formation of the tetrahedral intermediate if expulsion of the leaving group from [36] occurs faster than formation of [36]. Evidence to show the presence of the enamine [36] can be obtained by increasing its rate of formation and decreasing its rate of decomposition (Agathocleous et a/., 1985). There are several experimental observations which indicate that the reaction is not concerted and that expulsion of the leaving group at C(3’) occurs ofter p-lactam ring opening. The second-order rate constants for the hydroxide-ion catalysed hydrolysis of cephalosporins are correlated with 0,for C(3) substituents and give a Hammett p, of 2.5 for CH,L and of 1.35 for L (Fig. 8; Table 5). Several substituents at C(3), e.g. CH,, H, CH,CO,Et, are not expelled during hydrolysis or cannot be expelled directly by a concerted mechanism, e.g. CI. Substituents which are and those which are not expelled are controlled by the same linear free energy relationship (Proctor et a/., 1982; Page 1984a; Indelicato et al., 1974; Bundgaard, 1975). Leaving groups of different nucleofugalities (Stirling, 1979) influence the rate of reaction only by their inductive effect (Proctor el a/., 1982; Bundgaard, 1975). A series of cephalosporins with substituted pyridines and thiol leaving groups at C(3’) covering a range of 10 pK, units show a Brmsted PI, of 0.1 (Buckwell and Page, 1985). There is little or no change in the effective charge on the leaving group on going from the ground to the transition state. Esterification of the C(4) carboxylate group will make the P-lactam carbonyl carbon more electrophilic and facilitate p-lactam C-N bond

I

00

0*1

0.2

0-3

0-4

0.5

0-6

cI FIG. 8 Hammett plot of the second-order rate constants for the hydroxide-ion catalysed hydrolysis of cephalosporins [ 1 I] against Charton's cr, constant for the C( 3) su bsti tuen t .


205

cleavage but decrease the rate of fission of the C(3'h-L bond. The carbomethoxy group at C(4) induces biphasic kinetic behaviour in the alkaline hydrolysis of cephalosporins (Agathocleous et al., 1985). The two consecutive reactions observed spectroscopically are compatible; a stepwise mechanism for hydrolysis-fission of the p-lactam C-N bond to give the enamine [36] followed by expulsion of the leaving group to give [37]. TABLE5 Second-order rate constants (koH)at 30°C for the hydroxide-ion catalysed hydrolysis of cephalosporins as a function of substituents at C(3); I = 1.0 M(KC1)

I

co; X H CH3

c1

CH2S(CH2)3CH3 CH SCOCH CH20COCH3 CH,OH CH,~,H,

k , , / M - 's- ' 3.86 2.90 3.90 2.65 3.30 9.36 6.51 6.49

x x x x

low2 lo-'

x x

x x

lo-'

It is conceivable that the conversion of the enamine [36] to the a,p -unsaturated imine [37] is reversible. The addition of thiolate anions to the hydrolysis products of cephalosporins generated using P-lactamase as a catalyst indicates that the enamine and a$ -unsaturated imine are in equilibrium (Buckwell and Page, 1985). (See Section 12.) There is spectroscopic and kinetic evidence that the aminolysis of cephalosporins proceeds by a stepwise mechanism (see Section 12), and, in general, it appears that 3'-eliminations are not concerted with p-lactam C-N bond cleavage when cephalosporins react with nucleophiles. It has been frequently suggested (Flynn, 1972; Nishikawa and Tori, 1984) that the rate of basic hydrolysis of the p-lactam is correlated with the antibacterial activity of the antibiotic. Consequently, there have been many attempts to investigate possible correlations between structural parameters and chemical and biological reactivity. Biological activity is, of course, also very dependent upon the ability of the compound to penetrate into bacteria. There is a linear correlation between the logarithm of the rate constant for

206

MICHAEL I

PAGE

alkaline hydrolysis and the infrared carbonyl stretching frequency (Morin et al., 1969; Indelicato et al., 1974; Takasuka et al., 1982). Although the correlation has been claimed to be poor (Boyd, 1982), an extensive and careful examination (Nishikawa and Tori, 1981 ) supports the earlier claims. The higher the p-lactam vc=o value the greater is the p-lactam reactivity. The ''C nmr chemical shifts of the C(8) p-lactam carbonyl carbon varies over only a very narrow range (Paschal et a/., 1978; Dereppe et al., 1978; Schanck et a/., 1979). However, there does appear to be a good linear relationship between the logarithms of the rate constants, koH, for the base catalysed hydrolysis of cephalosporins and the differences between the 3C chemical shifts at C(3) and at C(4), A6(4-3) (Nishikawa and Tori, 1981, 1984; Mondelli and Ventura 1977; Schanck et a[., 1983). In addition to attempts to correlate hydrolysis rates with p-lactam C-N and C=O bond lengths and the degree of p-lactam nitrogen pyramidalisation (Sweet and Dahl, 1970), there have also been reports of the relationship between chemical reactivity and theoretically calculated parameters such as the net atomic charge on the 0-lactam carbonyl oxygen, the overlap population of the p-lactam carbonyl and transition state energies (Indelicato et al., 1974; Boyd, 1983; Boyd et a/., 1980; Petrongolo et al., 1980). In addition to the leaving group at C(3'), many other structural parameters within cephalosporins have been varied. Although the change from a A3- to a A2-cephem system causes the p-lactam nitrogen to become planar there is little difference, only 2- to 3-fold, in the chemical reactivity (Proctor et al., 1982). Other effects of structural changes are given in Table 4. Replacement of the dihydrothiazine S by 0 increases the rate of alkaline hydrolysis about 6-fold (Narisada et al., 1983). 1-Oxacephems sometimes show greater antibacterial activity than the corresponding cephalosporins (Firestone et al., 1977; Narisada et al., 1977; Murakami et al., 1981). Replacement of the ring S by CH, increases the rate of hydrolysis about 3-fold. It is interesting to note that this increase in reactivity is accompanied by a decrease in the p-lactam carbonyl stretching frequency which is contrary to the correlation described earlier (Nishikawa et al., 1982). Increased antibacterial activity of 1 -oxacephalosporins may result from a higher rate of penetration through the bacterial cell membrane because of increased hydrophilicity (Murakami and Yoshida, 1982). The addition of a methyl, methoxy or thiomethyl group at the 6-aposition of penicillin results in a reduction in antibacterial activity (Ho et d., 1973). In contrast. the addition of a 7-a-methoxy group to a cephalosporin results in compounds that are better transpeptidase enzyme inhibitors although they do not necessarily show better antibacterial properties (Indelicato and Wilham, 1974).


207

The introduction of a 7-a-methoxy group has an almost insignificant effect, less than 2-fold, upon the susceptibility of cephalosporins to alkaline hydrolysis (Indelicato and Wilham, 1974; Nishikawa et af., 1982; Narisada et a/., 1983). Inductively a 7-a-methoxy group should increase the rate 3-fold (Proctor et al., 1982) but unfavourable steric interactions in the tetrahedral intermediate must lower the rate. This steric effect is supported by the effect of a 7-a-methyl group which decreases the rate by a factor of 9 in cephalosporins (Narisada et a f . , 1983) whereas a 6-a-methyl decreases k,, for penicillins by a factor of 10 (Indelicato and Wilham, 1974). It is worth noting that 7-a-methoxycephalosporins have a less pyramidal p-lactam nitrogen (Applegate et al., 1974) and so its minimal effect upon chemical reactivity again indicates that the degree of pyramidalisation of nitrogen is not a major influence. Substituent changes in the 7-P-acylamido side chain have little effect upon chemical reactivity and yet can enormously change biological activity. For example, the incorporation of a syn-oxime function, as in cefuroxime, confers both high antibacterial activity and p-lactamase resistance (Bucourt et al., 1978; Schrinner et al., 1980). The oxime substituent, irrespective of its configuration, does not affect k,, for alkaline hydrolysis of the p-lactam but is highly enzyme specific. For example, the syn-isomer is 35-fold less reactive, as measured by kJK,,,, towards p-lactamase, whereas, the anti-isomer is twice as reactive compared with an analogous cephalosporin lacking the oxime function (Laurent et al., 1984). As the p-lactam ring of cephalosporins generally has a reactivity comparable with that of ethyl acetate, it is not surprising that hydrolysis of an acetoxy ester side chain at C(3) is competitive with hydrolysis of the P-lactam. The second-order rate constant for the base-catalysed conversion of the ester to the 3-hydroxylmethylcephalosporin is usually similar to that for p-lactam hydrolysis (Yamana and Tsuji, 1976; Berge et al., 1983; Bundgaard, 1975). 5 Acid hydrolysis

The pH-rate profile for the hydrolysis of benzylpenicillin is shown in Fig. 9 (Gensmantel et a f . ,1978). There is no significant spontaneous hydrolysis but the p-lactam does undergo an acid catalysed degradation. Also shown in Fig. 9 is the pH-rate profile for the hydrolysis of cephaloridine. There are two immediate differences: cephalosporins exhibit a spontaneous pHindependent hydrolysis and are less reactive towards acid than penicillins by a factor of about lo4 (Proctor er af., 1982). In addition to the expected hydrolysis product, benzylpenicilloic acid [30], the acid catalysed degradation of benzylpenicillin gives benzylpenicillenic

Log k IS-' 10 obs

\

- 2.c

\ \ t

\ +,

'?

- 3.c

- 4.0

-54

- 64

-7-C

0

2

4

6

8

10

12

14

PH FIG.9 pH-Rate profiles for the hydrolysis of benzylpenicillin ( + ) and cephaloridinc ( 0 )at 30°C; I = 1.0 M(KC1)


209

acid [38], benzylpenamaldic acid [39], benzylpenillic acid [40] and benzylpenilloic acid [41]. The proportion of each product formed depends upon the pH (Schwartz, 1965; Dennen and Davis, 1962). Although several kinetic studies have been reported on the degradation of penicillins in acidic media (Schwartz, 1965; Longridge and Timms, 1971; Degelaen et al., 1979a; Blaha ei al., 1976; Bundgaard, 1980; Kessler et al., 1983) there is still considerable uncertainty about the details of the reaction pathway.

[401

[411

It has been suggested (Schwartz, 1965) that one reaction pathway involves specific acid catalysed hydrolysis of the unionised penicillin to give penicilloic acid [30] and, by decarboxylation of this product, penilloic acid [40]. In a second, parallel reaction penicillin rearranges to penicillenic acid [38] which subsequently degrades to penillic acid [40]. The formation of penicillenic acid was suggested to be the result of a specific acid catalysed reaction of ionised penicillin or the kinetically equivalent spontaneous rearrangement of the unionised acid. Another study (Blaha et al., 1976) proposed that penicillenic acid is the first formed product and that all other degradation products are formed from this. Based on an nmr spectroscopic study and the extent of deuteriation at C(6), it was concluded that three parallel degradation reactions occur and penicillin initially gives directly penicilloic acid, penillic acid and penicillenic acid. In addition to these detectable intermediates it has also been proposed (Bundgaard, 1980; Proctor et al., 1982) that an oxazolone-thiazolidine intermediate [42] is formed which is the precursor of the degradation products. The proposal (Degelaen et al., 1979a) that penillic acid is formed directly from penicillin has been criticised and the optimisation of fitting kinetic data to the nmr observations suggests that all the penicillenic acid is transformed

210

MICHAEL I. PAGE

into penamaldic acid [39] (Kessler et al., 1983). Furthermore, it was suggested that penillic acid is the major degradation product of penicilloic acid. A study of substituent effects and the acidity dependence of the rate of degradation shows that most of these problems can be resolved into an acceptable reaction pathway (Proctor et al., 1982). Thermodynamically the most basic site for the protonation of normal amides is oxygen and the pK, of 0-protonated amides is 0 to -3, (Liler, 1969; Yates and Stevens, 1965), although it is claimed that N-protonation occurs in dilute acid and O-protonation in strong acid (Liler 1975). Whether the mechanism of the acid catalysed hydrolysis of simple amides proceeds via 0- or N- protonation seems to have been resolved in favour of the former (Williams 1976; Modro et al., 1977; McLelland and Reynolds, 1974; Kresge et al., 1974), although there remain some dissenting voices (Liler, 1975). Remarkably, the logarithms of the pseudo first-order rate constants for the hydrolysis of some p-lactam antibiotics and derivatives increase linearly with decreasing H , values up to -5 (Proctor ef af., 1982). This is quite unlike the behaviour of other amides for which the rate of hydrolysis passes through a maximum, attributed to complete conversion of the amide into its 0-conjugate.acid and to decreasing water activity (O’Connor, 1970; Smith and Yates, 1972). This indicates that the p-lactams are far less basic than normal amides for 0-protonation and that a different mechanism of hydrolysis is operating. Neither the nitrogen nor the oxygen of the bicyclic p-lactams is sufficiently basic for substantial conversion to the conjugate acid; pK, for 0- or N- protonation must be < -5. This behaviour is not peculiar to bicyclic p-lactams since monocyclic p-lactams show similar reactivities and behaviour (Proctor et al., 1982; Wan et al., 1980). The reduced basicity of cyclobutanones compared with other ketones has been previously noted (McLelland and Reynolds, 1976) and the very weak basicity of p-lactams may have a similar origin (Bouchoux and Houriet, 1984). The slopes of plots of the logarithms of the pseudo first-order rate constants against H , are - I to - 1.3 and, since water activity decreases with increasing acidity, it appears that water is not involved in the transition state. This can be explained by a unimolecular A - I type mechanism with N-protonation of the p-lactam (Scheme 5). That N-protonation takes place

MECHANISMS OF REACTIONS OF B-LACTAM ANTIBIOTICS

21 1

H

Scheme 5

is not the result of reduced amide resonance in penicillins and cephalosporins but must be an intrinsic property of p-lactams. The introduction of the A-I mechanism could result because the normal A-2 mechanism is retarded or because the A-1 pathway is favoured. The most likely explanation is the enhanced rate of C-N bond fission that occurs in p-lactams as a result of the relief of ring strain (Proctor et al., 1982). Substituents at C(6) in penicillins aflect both the carbonyl carbon and nitrogen of the p-lactam inductively, but their effect on C-N bond cleavage will be that predominantly of an acyl substituent. As described earlier, penicillins with an acylamido side chain at C(6) undergo degradation in acid solution to give a variety of products which must result from transformations involving the side chain amide. Electron-withdrawing substituents which cannot be involved in neighbouring group participation greatly retard the rate with a Hammett p,-value of ca -4.0 to -5.0, depending upon the acidity (Proctor et al., 1982). The effect of acyl substituents upon the rate of acid catalysed hydrolysis of acyclic amides is small, with electron-withdrawing substituents producing either a small increase or decrease in rate (Bruylants and Kezdy, 1960; Bolton and Jackson, 1969).

21 2

MICHAEL I. PAGE

Electron-withdrawing substituents in the amine portion of the p-lactam decrease the rate of acid catalysed degradation of penicillins. The Br~nsted p-value is ca 0.35 (Proctor et al., 1982) compared with -0.26 for acyclic anilides and amides (Giffney and O'Connor, 1975). Although the effects of substituents are not large, they are significant and in the opposite direction for p-lactams compared with other amides which again is indicative of a different mechanism. There is a large dependence of the rate of the acid catalysed degradation of penicillins upon the nature of the acylamido side chain. Electron-withdrawing substituents decrease the rate of degradation and for substituted phenylpenicillins the Hammett p-value is - 1.60 (Yamana et al., 1974b). Because of the similarity of the rate constants for the degradation of benzylpenicillin and its methyl ester there is no evidence for neighbouring group participation of the carboxy-group in the fission of the p-lactam ring (Proctor et al., 1982). However, penicillenic acid [37] is not formed significantly below pH I and is the major product only at ca pH 3 (Schwartz, 1965; Blaha et al., 1976; Bundgaard, 1980). It has been suggested (Schwartz, 1965) that penicillenic acid is formed from the specific acid-catalysed reaction of ionised penicillin or the kinetically equivalent spontaneous rearrangement of the undissociated acid. This suggestion is based on the observation that the rate of formation of penicillenic acid reaches a plateau at pH ca 2, where the overall rate constant for degradation also shows an inflection point, thought to correspond to ionisation of the carboxy-group. However, benzylpenicillin methyl ester shows exactly the same behaviour and the rate of penicillenic acid formation follows (2) with k , = 3.42 x 10-4s-' and K, = 10-2.25 (Proctor et a/., 1982). The ionisation of a group with pK, 2.25 obviously cannot correspond to the carboxy-group as suggested for benzylpenicillin itself. It seems doubtful, therefore, that the decreased formation of penicillenic acid from penicillin at low pH is due to protonation of the ionised carboxy-group.

The acid catalysed degradation of C(6) acylamido penicillins show a rate of enhancement of ca lo3 compared with that predicted from the Hammett plot for C(6) substituents (Proctor et al., 1982). The mechanism of degradation must therefore incorporate the acylamido group in the rate-limiting step or in a pre-equilibrium step. The important facts to be explained therefore are: 1. The rate enhancement observed for the acid catalysed degradation of benzylpenicillin indicates neighbouring group participation by the acylamido side chain.


21 3

2. Benzylpenicillenic acid [38] is not formed in acidic solution (pH < 1 ) where its rate of formation is pH-independent. 3. The majority of the penillic acid [40] and penicilloic acid [30] formed does not come from a mechanism involving D-incorporation at C(6), i.e. from penicillenic acid type intermediates (Degelaen et af., 1979a; Kessler et al., 1983). 4. The rate-limiting step for the acid catalysed degradation of all penicillins does not appear to involve water. A mechanism compatible with these observations is shown in Scheme 5 (Proctor et al., 1982). Reversible ring opening of the p-lactam gives an acylium ion which may be trapped by the intramolecular amido group rather than by water to give the protonated oxazolone-thiazolidine [42]. This can react with water to give penicilloic acid, undergo intramolecular nucleophilic attack of the thiazolidine nitrogen on the carbon of the protonated oxazolone to give penillic acid [40] or eliminate across C(5)-C(6) to give penicillenic acid [38]. The pKa-value of the protonated oxazolone or that of the thiazolidine nitrogen could be the kinetically important one controlling penicillenic acid formation. The protonated oxazolone probably has a pKa of ca 0 and furthermore there is no obvious chemical reason why protonation of the imine should inhibit elimination. For the thiazolidine nitrogen pK, is estimated to be ca 3. Protonation of this nitrogen would inhibit penicillenic acid formation. The kinetically important ionisation is therefore attributed to that of the thiazolidine nitrogen. There is a slight shadow cast over the neat picture for the acid catalysed degradation of p-lactams outlined in Scheme 5. The rate of penicillin degradation is ca lo3 faster than that predicted from the o-value for RCONH. The implication is that the acylamido-group participates in the rate-limiting step, which is acceptable if k , is rate-limiting (Scheme 5) for hydrolysis, i.e. the formation of the acylium ion is reversible. However, this then reintroduces water in the transition state for hydrolysis of the penicillins lacking an acylamido-side chain, and this is not indicated by the acidity dependence of the rate of reaction. An alternative explanation is that attack of the acylamido-group on the p-lactam carbonyl carbon is concerted with C-N bond fission, i.e. the reaction does not proceed via the acylium ion. The acid hydrolysis of cephalosporins shows similar behaviour to that of the penicillins, but they are about 104-fold less reactive (Proctor et al., 1982). Electron-withdrawing substituents at C(7) in cephalosporins decrease the rate of acid hydrolysis and, as for penicillins, the Hammett p,-value is ca -5. There is no evidence for neighbouring group participation by the 7-acylamido-group as postulated for the penicillins. There seems no obvious explanation of the difference in behaviour between the cephalosporins and penicillins. Either attack of water on the acylium ion (Scheme 5) could be

MICHAEL I PAGE

21 4

inhibited in penicillins relative to cephalosporins, and therefore trapping of the acylium ion by the acylamido-group is more effective in penicillins, or the latter process could be inhibited in cephalosporins. One difference is that the dihydrothiazine of cephalosporins has a less basic nitrogen than the thiazolidine of penicillins; enamine resonance lowers the pK,-value by ca 2 units. This could account for the lower reactivity of 7-amino- and 7-chlorocephalosporanic acid compared with the analogous penicillin derivatives. Similar to alkaline hydrolysis there is no evidence for the group at C(3') in cephalosporins (acetate or pyridine) affecting the rate of reaction. In fact, the 3-methyl derivative is more reactive than the cephalosporins with acetate or pyridine at C(3') which yet again indicates that expulsion of these groups is not important in the rate-limiting step. At pHs above the pK, of the 3-carboxyl group, the rate of degradation of penicillin is acid catalysed up to about pH 6. The pathway for hydrolysis could be an acid catalysed reaction of the penicillin with an ionised carboxylate or the kinetically equivalent mechanism, the spontaneous degradation of penicillin with an unionised carboxyl function as indicated in (3). k,[PenCO;][H

'1

= k,K,[PenCO,H] =

k , [PenCO,H]

(3)

The calculated rate constant k , for benzylpenicillin at 30°C (Proctor et al., 1982) is 6.30 x 10-4s-' which is about 6000 times greater than the spontaneous hydrolysis of benzylpenicillin with an ionised carboxylate at C(3). Some investigators have favoured the spontaneous degradation of unionised penicillin mainly because of the lack of an appreciable ionic strength effect on k , (Finholt et al., 1965; Yamana et al., 1974b) and an observed solvent isotope effect k,H20/k,DZo of 1.53 at 60°C (Yamana et al., 1977). Penicillenic acid and methyl benzylpenicillenate are formed during the degradation of benzylpenicillin and benzylpenicillin methyl ester, respectively (Proctor et al., 1982; Bundgaard, 1980). The mechanism of penicillin degradation could involve intramolecular participation by the C(3) carboxyl group or the 6-amido side chain. The second-order rate constant k , for the catalysed degradation of penicillin with an ionised carboxyl at C(3) is 0.354 M - ' s - ' at 30°C compared with 8.2 x lo-' M - ' s - ' for benzylpenicillin methyl ester. This rate difference of 4-fold is insignificant. The calculated first-order rate constant k , for undissociated benzylpenicillin degradation is 10-fold greater than that for the methyl ester of benzylpenicillin. This is also probably too small to be indicative of intramolecular catalysis by the carboxyl group. The 6-amido side chain also does not provide a significant rate enhancement in the degradation of penicillins between pH 3 and 6. The second-order rate constants for the acid catalysed degradation of 6-ammonium penicilla-

MECHANISMS OF REACTIONS OF O-LACTAM ANTIBIOTICS

21 5

nate and 6-aminopenicillanate are 3.5 x and 0.675 M - ' s - l , respectively, compared with 0.354 M - ' s - ' for benzylpenicillin at 30°C (Proctor et al., 1982). Replacement of the 6-amido side-chain by an amino group therefore leads to a slight rate enhancement. It appears that the amido side chain is not involved in the rate-limiting step of the degradation of penicillins from p H 3 to 6, but the mechanism of this degradation is still ambiguous. 6 Spontaneous hydrolysis

Penicillins undergo an acid and a base catalysed hydrolysis but there is no significant uncatalysed reaction. The pH minimum is at 7 (Fig. 9) and k,, the apparent first-order rate constant for spontaneous or water-catalysed degradation, is 1 x 10-7s-' at 30°C (Gensmantel et al., 1978). By contrast, cephalosporins show a pH independent reaction between pH 3 and 7 with k , in the range 5 x to 3 x 10-6s-' at 30°C (Yamana and Tsuji, 1976; Yamana et al., 1974a; Fujita and Koshiro, 1984; Berge et al., 1983; Lumbreras et al., 1982). It has been suggested that this pHindependent reaction involves intramolecular nucleophilic attack on the p-lactam by the 7-amido side chain (Yamana and Tsuji, 1976). There are several problems raised by this proposal. As outlined in the previous section, cephalosporins do not show neighbouring group participation by the 7-amido side chain in their acid catalysed degradation. It is difficult to understand why it would therefore occur in the uncatalysed reaction. Furthermore, 7-amido cephalosporins show a similar reactivity to 7-aminocephalosporanic acid in their spontaneous degradation. However, the deuterium solvent isotope effect k,H20/k,D20 is 0.93 at 35°C (Yamana and Tsuji, 1976) which is not typical of a water catalysed hydrolysis. Although cephaloridine [pyridine at C(3')] is 6-fold more reactive towards hydroxide ion than cephalothin [acetate at C(3')], it is 3-fold less reactive in its spontaneous degradation (Proctor et al., 1982; Yamana and Tsuji, 1976). Broad plateaus around neutral pH are observed for the hydrolysis of some penicillins, e.g. cloxacillin (Bundgaard and Ilver, 1970), phenethicillin (Schwartz et al., 1962), cyclacillin (Yamana et al., 1974b) and dicloxacillin (Pawelczyk et al., 1981). The rate constants for the spontaneous hydrolysis of most penicillins are ca lO-'s-' at 30°C and there is little dependence of the rate upon the nature of the side chain (Yamana et al., 1974b). Unlike the spontaneous degradation of cephalosporins, that of penicillins shows a significant solvent isotope effect, k,H20/k,D20 is 4.5 at 60°C (Yamana et al., 1974b). The most likely mechanism of spontaneous hydrolysis therefore involves general base catalysis by water of nucleophilic attack by water on the p-lactam as indicated in [43], although a rate-limiting step involving breakdown of the intermediate is also possible.

21 6

MICHAEL I. PAGE

H'

0 H '

7 Buffer catalysed hydrolysis

The hydrolysis of both penicillins and cephalosporins are often catalysed by buffers. Catalysis by borate and phosphate buffers is usually found for the hydrolysis of some cephalosporins, but buffer catalytic effects are not always observed. Sometimes catalysis by phosphate monoanion is kinetically more important than that by the dianion (Lumbreras et al., 1982). TABLE 6 Summary of second-order rate constants for the reaction of benzylpenicillin with oxygen anions at 30°C; I = 1.0 M(KC1) Conjugate acid of anion

PK,

kR0-/M-'s-'

H2O CH,OH EtOCH,CH,OH CICH,CH,OH CH ECH-CH~OH CF,CH,OH (CF,),CHOH HCO; (CF,),C(OH), H,PO; CH,CO,H HC0,H

13.83 15.54 14.98 14.31 13.55 12.43 9.30 9.71 6.76 6.51 4.60 3.62

1.54 x l o - ' 15.96 3.45 20.1 7.82 x lo-' 2.23 x lo-' 1.05 x 1.25 x 8.98 x 1.08 x 1 0 - 5

1.26 x 4.47 x lo-'

Rate constants for the hydrolysis of benzylpenicillin catalysed by oxygen bases are given in Table 6 . The corresponding Brernsted plot is shown in Fig. 10; its non-linearity is indicative of a change in mechanism. The Brernsted p-value for weak bases is 0.39 and probably represents general base catalysed hydrolysis. The reaction with alkoxide ions is nucleophilic. The

Log k M-lS1 10 RO' 2.0 CLCH2CH20-+ 1.0

/

I

0.0

- 19 - 2.0 - 3.0 - 4.0 2-

HPO,

- 5.0

I

Hcwco-

- 6.0 - 7.0 2

4

6

8

10

12

14

PIC

FIG. 10 Brransted plot of the oxygen-anion catalysed hydrolysis of benzylpenicillin at 30°C. The steep slope represents nucleophilic catalysis whilst the less steep one for weakly basic anions represents general base catalysis

21 8

MICHAEL I

PAGE

Brcansted P for bases whose conjugate acids have pK, > 7 is 0.95 and represents nucleophilic catalysed hydrolysis (Proctor and Page, 1979). Evidence for an intermediate ester formed during the reaction has been obtained with alkoxide ions and phosphate dianion (Proctor and Page, 1979; Bundgaard and Hansen, 1981). Weakly basic catalysts probably act as general base catalysts. The reaction with acetate exhibits a solvent isotope effect k,H20/k,D20 of 2. I , as expected for general base catalysis but not for a nucleophilic reaction (Anderson et al., 1961). The reaction of benzylpenicillin with imidazole is dominated by imidazole buffer catalysis and hence the rate term which is first order in imidazole is difficult to determine (Bundgaard, 1976a, 1972). The rate constant reported for imidazole is similar to that for phosphate dianion which is surprising if they both represent nucleophilic catalysed hydrolysis. The mechanism of the reaction with alkoxide ions is discussed in Section 13 but it can be noted here that the nucleophilic reaction probably proceeds by rate-limiting breakdown of the tetrahedral intermediate [44]. Fission of the J3-lactam C-N bond may generate the amine anion or be catalysed by water acting as a general acid. Benzylpenicillin is about 1000-fold less reactive towards oxygen nucleophiles than is acetylimidazole which involves rate-limiting expulsion of the imidazole anion (Oakenfull and Jencks, 1971).

dR

8

Metal-ion catalysed hydrolysis

Transition-metal ions cause an enormous increase in the rate of hydrolysis of penicillins and cephalosporins (Gensmantel et al., 1978, 1980; Cressman et al., 1969). For example, copper(1I) ions can enhance the rate of hydrolysis of benzylpenicillin 108-fold, a change in the half-life from 11 weeks to 0.1 seconds at pH 7. In the presence of excess metal ions, the observed apparent first-order rate constants for the hydrolysis of the P-lactam derivatives are first order in hydroxide ion but show a saturation phenomenon with respect to the concentration of metal ion which is indicative of the formation of an antibiotic/metal ion complex. A kinetic scheme is shown in (3), where M is


21 9

the metal ion and L is the p-lactam, and some relevant data are given in Tables 7 and 8. The rate of hydroxide-ion catalysed hydrolysis of benzylpenicillin bound to metal ion shows the following rate enhancements compared with the uncoordinated substrate: Cu(II), 8 x 10’; Zn(II), 4 x lo5; Ni(II), 4 x lo4; Co(II), 3 x lo4. The analogous data for cephaloridine are: Cu(II), 3 x lo4, Zn(II), 2 x lo3 (Gensmantel et al., 1980). TABLE7 Summary of the rate and association constants for the copper(I1)-catalysed hydrolysis of p-lactam derivatives in water at 30°C ( I = 0.5 M)”

Benzylpenicillin Benzylpenicillin methyl ester 6-p-Aminopenicillanic acid Penicillanic acid Cephaloridine 3-Methyl-7o-phenylacetamido ceph-3em-4-carboxylic acid

8 x lo7

0.154

187

1.22 x lo7

2.51

-

< 3 x 104“ < I . S x 104

6.35 x lo-’

232

5.15 x lo6

8 x lo7

7.40 x 1 0 - 3 0.526

120 2080

1.58 x 106 1.64 x 104

2 x 107 3 x 104

2.41 x lo-’

2400

1.56 x 103

7 x 104

Gensmantel ef al., 1980 Second-order rate constant for the hydroxide-ion catalysed hydrolysis Association constant for metal ion and p-lactam Second-order rate constant for the hydroxide-ion catalysed hydrolysis of metal-ion bound p-lactam k,O”K, saturation was not observed.

COORDINATION SITE

Copper(I1) ion coordinates to the carboxylate group and the p-lactam nitrogen of benzylpenicillin as shown in [26] (Gensmantel et al., 1980). Coordination occurring to the carboxylate group is indicated because esterification of this group decreases the rate enhancement by a factor of ca 5 x lo3. Nonetheless, the rate of the hydroxide-ion catalysed hydrolysis of the copper(I1)-bound methyl ester of benzylpenicillin is still 1.5 x 104-fold faster than the rate of hydrolysis of the uncoordinated ester, although the product of the reaction is not known.

MICHAEL I. PAGE

220

TABLE 8

Summary of the rate and association constants for the metal-ion catalysed hydrolysis of benzylpenicillin and cephaloridine in water at 30°C ( I = 0.5 M)"

CU(11)

Zn(11) Ni( 11) Co(I I)

1.22 x 10' 6.0 x 104 5.9 x 103 4 . 1 x 103

187 109 1 I9 178

1.64 x 104 8.75 x 10'

2080 2181

'Gensmantel et al., 1980 ' Second-order rate constant for the hydroxide-ion catalysed hydrolysis Association constant for metal ion and p-lactam It has been suggested that copper(I1) ions coordinate to the 6-acylamino side chain and the p-lactam carbonyl group (Cressman et af., 1969). However, replacement of the acylamino side chain by the more basic amino group has little effect upon the binding constant and the rate enhancement for the hydroxide-ion catalysed hydrolysis for 6-aminopenicillanic acid [ 101 is very similar to that for benzylpenicillin. Furthermore, complete removal of the amido side chain as in penicillanic acid, also gives similar binding constants and rate enhancements. It is apparent from these observations that copper(I1) ions do not bind to the amido side chain in penicillins, and that coordination probably occurs between the carboxylate oxygen and the p-lactam nitrogen [26] (Gensmantel et a/., 1980; Fazakerley et al., 1976; Fazakerley and Jackson, 1977). Model studies of the binding of metal ions to a-amido carboxylic acids indicate that coordination between the carboxylate group and the amide nitrogen does occur although the evidence is not unambiguous (Fazakerley and Jackson, 1975a,b; Weiss et a/., 1957; Weiss and Fallab, 1960; Eichelberger et a/., 1974; Ishidate ef a/., 1960; Taguchi, 1960; Eichorn, 1973). Copper(I1) ions bind 10-fold more tightly to cephalosporins than to penicillins which would be surprising if the sites of coordination were similar. Molecular models indicate that one of the conformations of cephalosporins would be very suitable for metal-ion coordination between the carboxylate group and the p-lactam carbonyl oxygen [45]. The shortest distance between the carboxylate oxygen and the p-lactam nitrogen is similar ( 2.7 A) in penicillins and cephalosporins. However, the carboxylate oxygen-p-lactam oxygen shortest distance is much smaller ( 2.7 A) in cephalosporins than in penicillins ( -4.6 A). Precipitation of the p-lactam/metalion complex in the presence of excess ligand gives solids with interestingly different characteristics. Benzylpenicillin forms a 1 : 1 complex with both

-

-

MECHANISMS OF REACTIONS

OF

8 - L A C T A M ANTIBIOTICS

221

RCONH

F fs'l 1451

copper( 11) and zinc(l1) in which the asymmetric stretching frequencies of the p-lactam carbonyl and the carboxylate are decreased by ca 30cm-' compared with uncoordinated penicillin. The nmr spectrum of the zinc(II)/benzylpenicillin complex shows a downfield shift for the C(3) hydrogen, consistent with the proposed mode of binding (Gensmantel et al., 1981; Fazakerley and Jackson 1975a,b; Asso et al., 1984). However, solid ML, complexes of Mn(II), Pb(I1) and penicillins have been reported (Chakrawarti C I al., 1982, 1984). Alkali-metal salts affect the infrared spectra of penicillin; the p-lactam carbonyl stretch is split in the potassium salt (1777 and 1757cm-'), and the amide I band is at 1669cm-' for the potassium salt and at 1700cm-' for the sodium salt (Zugara and Hidalgo, 1965). Thorium(1V) forms a complex with benzylpenicillin but not with its methyl ester (Ishidate et al., 1960). Cephalothin and 3-methyl-7~-phenylacetamidoceph-3-em-4-carboxylic acid form solid 2 : 1 complexes with transition-metal ions in which not only is the asymmetric stretching frequency of the carboxylate decreased, but also the p-lactam carbonyl stretching frequency by 10-30 cm- ', depending upon the nature of the metal ion. The nmr spectrum of the zinc(II)/cephalothin complex shows a downfield shift of the C(7) hydrogen. The site of metal-ion coordination could thus be different for cephalosporins and involve the p-lactam carbonyl oxygen, although the situation in solution may be different. For example, the kinetic data indicate only a 1 : 1 complex and, of course, the thermodynamically favoured binding site is not necessarily the kinetically important one. RATE ENHANCEMENT

The hydroxide-ion catalysed hydrolysis of benzylpenicillin probably proceeds by the formation of the tetrahedral intermediate [46]. The pK,-value of the bridgehead nitrogen in [46] is estimated to be 8.0 (Page and Jencks, 1972b) so there is an enormous change in the basicity ( > 12pK, units) of the p-lactam nitrogen as the reaction proceeds. The role of the metal ion in the hydroxide-ion catalysed hydrolysis is to stabilise the tetrahedral intermediate. An estimate of the binding constant of copper(I1) ions to [46] can be made from a comparison with model compounds. Based on a Brnnsted plot

MICHAEL I. PAGE

222

for the binding of model ligands such as thioproline, proline and glycine, the estimated association constant for copper(I1) and [46] is lo7.' M - ' (Gensmantel et al., 1980).

[461

The rate of the hydroxide ion catalysed hydrolysis of copper(I1)-bound benzylpenicillin is 8 x lo7 faster than that of uncoordinated benzylpenicillin (Gensmantel et al., 1978). A better estimate of the stabilisation of the transition state by the metal ion is from the comparison of the third-order rate constant, k , K , , for the metal and hydroxide-ion catalysed hydrolysis with the second-order rate constant for the hydroxide-ion catalysed hydrolysis. For copper(I1) ions and benzylpenicillin this ratio is 1.2 x 10" M. Copper(I1) ion thus stabilises the transition state for hydroxide ion catalysis by 13.9 kcal mol-' at 30°C compared with an estimated value of 9.8 kcal mol-' for the stabilisation of the tetrahedral intermediate [46]. EFFECT OF T R A N S I T I O N METAL ION

The rate enhancements brought about by various metal ions for the hydrolysis of penicillins and cephalosporins are summarised in Tables 7 and 8. There is no correlation between the binding constant of the p-lactam antibiotic with the metal ion and the rate enhancement. The order of reactivity is that of the Irving-Williams series (Irving and Williams, 1953): Co(I1) < Ni(I1) < Cu(I1) > Zn(I1). EFFECT OF

P-LACTAM

Copper(I1) ions bind ca 10-fold more tightly to cephalosporins than to penicillins (Table 7). This is at first surprising in view of the greater non-planarity of the penicillin molecule and the correspondingly greater basicity of the p-lactam nitrogen which is assumed. In cephalosporins the possibility of enamine type conjugation and the less favourable geometry [sp' C at C(3) and C(4)] for metal-ion coordination to the p-lactam nitrogen and the carboxylate group would be expected to hinder coordination. Nonetheless, the rate of hydroxide-ion catalysed hydrolysis of copper(I1)bound cephaloridine is ca 3 x 104-fold faster than that for the uncoordinated compound. This may be compared with a rate enhancement of

MECHANISMS OF REACTIONS OF p - L A C T A M ANTIBIOTICS

223

8 x lo7 for benzylpenicillin. The ratio of the third-order rate constant, k , K , , for the copper(I1) ion plus hydroxide ion catalysed hydrolysis of cephaloridine to the second-order rate constant for hydroxide-ion catalysed hydrolysis of the same substrate is 1.6 x lo8 M. The corresponding ratio for benzylpenicillin is 1.2 x 10" M. The transition state for cephaloridine hydrolysis is therefore stabilised by copper(I1) ions ca 100-fold less than that for penicillin hydrolysis, but both transition states are greatly stabilised by the metal ion. Again, ad hoc explanations for this difference may be found in the lower basicity of the ring nitrogen in the tetrahedral intermediate formed from cephaloridine and/or a less favourable geometry. Whether or not the group at C(3') of the cephalosporin [2] is expelled or not makes little difference to the rate enhancement brought about by the metal ion. The 3-methyl derivative, has a similar association constant for binding of copper(I1) ion to that for cephaloridine [ I I ] (X = CH,-pyridinium) (Table 7). The rate enhancement brought about by copper(I1) ion is the same within a factor of 2 (Gensmantel et al., 1980). It has been suggested that a ternary complex is formed between benzylpenicillin, zinc(I1) and tris buffers and that hydrolysis occurs by intramolecular nucleophilic attack of one of the coordinated buffer hydroxyl groups on the 0-lactam (Schwartz, 1982; Tomida and Schwartz, 1983). 9 Micelle catalysed hydrolysis of penicillins

Non-functionalised micelles which catalyse reactions provide a simple illustration of the utilisation of the binding energy between a phase or macromolecule and the reactants to lower the free energy of activation (Jencks, 1975; Page, 1980b; Fendler and Fendler, 1975; Bunton, 1977, 1984; Cordes, 1978; Bunton and Savelli, 1986). Even if the rate constant for a bimolecular reaction within the micelle is the same as that in the bulk solvent, a rate enhancement may be observed if the reactants are confined to a smaller volume within the micelle (Jencks 1975; Martinek et al., 1973, 1975). This requires the free energy of interaction between the reactants and the micelle to compensate for the loss of entropy resulting from the restriction of the reactants within the micelle. Micelles are of particular interest with respect to the hydrolysis of penicillins because they can provide different microenvironments for different parts of the reactant molecule. There is a nonpolar, hydrophobic core that can provide binding energy for similar groups on penicillin and a polar, usually charged, outer shell that can interact with the penicillin's polar groups. Hydrophobic substrates and counterions are attracted to the micelle so that a cationic micelle should assist the reaction between a neutral molecule and anionic nucleophile while anionic micelles will inhibit such reactions.

224

MICHAEL I. PAGE

The micelle catalysed hydrolysis of penicillins in alkaline solution is unusual because it involves the reaction between two anions, the hydroxide ion and the negatively charged benzylpenicillin (Gensmantel and Page, 1982a). The rate of the hydroxide-ion catalysed hydrolysis of benzylpenicillin decreases approximately three-fold in micellar solutions of itself (Hong and Kostenbauder, 1975). -4

2 x 1 0 M Pen.

L

I

10

3.0

SO

7.0

FIG. 1 1 Observed pseudo first-order rate constants for the hydrolysis of benzylpenicillin at the concentrations shown as a function of cetyltrimethylammonium bromide (CTAB) concentration at 30°C (Gensmantel and Page, 1982a)

The acid catalysed degradation of penicillins is inhibited in cationic micelles of cetyltrimethylammonium bromide (Tsuji et af., 1982) and, as expected, neither anionic micelles of sodium dodecylsulphate nor polyoxyethylene lauryl ether promote the hydroxide-ion catalysed hydrolysis of benzylpenicillin (Gensmantel and Page, 1982a). In the presence of cetyltrimethylammonium bromide (CTAB) the pseudo first-order rate constants for the alkaline hydrolysis increase rapidly with surfactant concentration once


225

above the critical micelle concentration (cmc) of the surfactant (Gensmantel and Page, 1982). Increasing the surfactant concentration eventually leads to a slow decrease in the observed rate (Fig. 11). This general shape of surfactant-rate profile has been found for many bimolecular reactions catalysed by cationic micelles (Menger and Portnoy, 1967; Yatsimirski ef al., 1971; Bunton et al., 1970). However, unusually, the observed pseudo first-order rate constant is not independent of penicillin concentration. The binding constant between the micelle and substrate is unlikely to change significantly with concentration and yet the lower the concentration of benzylpenicillin the faster the rate increases and the greater the maximal rate obtained, the rate maximum shifting to a lower surfactant concentration. This observation could be explained if both hydroxide ion and benzylpenicillin compete for the same types of sites in the micelle and if benzylpenicillin binds better than hydroxide ion. Increasing the hydroxide ion concentration inhibits the rate of the micellar catalysed reaction while the rate in the bulk aqueous phase increases. The observed pseudo first-order rate constant for the micelle catalysed hydrolysis does not increase linearly with increasing hydroxide-ion concentration at constant surfactant concentration but reaches a maximum value (Gensmantel and Page, 1982a). The kinetic evidence implies that there must be some binding between the benzylpenicillin anion and the micelles of CTAB and this has been shown spectroscopically (Chaimovich el al., 1985). The maximum rate acceleration in the alkaline hydrolysis of benzylpenicillin by CTAB micelles is about 50. The rate increase observed for many reactions upon the addition of detergents above the cmc has been explained on the basis of Scheme 6 K.

S f M

MS

I

1'-

P

'"'

P Scheme 6

(Menger and Portnoy, 1967). The substrate, S, associates with the micelle, M, to form a substrate-micelle complex MS with an equilibrium constant K,. The substrate and substrate-micelle complex form the product P with rate constants k, and k , referring to bulk aqueous and micellar phases, respectively. The observed first-order constant is given by (4) in which C , is the concentration of micelles. Equation (4) has been used to explain catalysis

226

MICHAEL I. PAGE

in the presence of surfactant and, since it takes a similar form to the Michaelis-Menten equation for enzyme catalysed reactions, the rate and equilibrium constants can be evaluated by the procedure of double reciprocal plotting (Martinek et al., 1977). However, (4) cannot explain many experimental observations for bimolecular or higher order reactions when two species or reactants compete for the same vacant “sites” in the micelles. In recent years development of the kinetic theory for micellar reactions of molecularity greater than one have led to two general approaches to micellar catalysis. Equation (5) has been derived (Martinek et al., 1973,

1975) for the reaction between two unchanged molecules and quantitatively explains surfactant-rate profiles for these bimolecular reactions; P A and P, are the partition coefficients of molecules A and B between aqueous and micellar phase, V is the molar volume of the micelle and C, is the concentration of surfactant. Romsted (1977) has suggested an expression, (6), which considers the effects of ions present in solution as well as substrate binding to the micelle.

The major assumption in deriving (6) is that the total number of counterions bound to a micelle is constant, allowing evaluation of the rate constant associated with the substrate-micelle complex. Exchange constant Ki iefers to equilibrium (7) where I is the reactive, and X the unreactive counterion,

(total concentrations C,(, C,,), and the subscripts m and w refer to the micelle and bulk phases, respectively. The constant p is the degree of binding of the counterions to the Stern layer, S is the molar density of the micellar phase and C , is the surfactant concentration. The inhibitory effect of increasing benzylpenicillin concentration can be rationalised by the pseudo-phase ion-exchange model, but as the number of molecules of the antibiotic bound to the micelle increases ( > 10 for concenM) the behaviour of a micelle covered with benzylpenitrations > 2 x cillin is probably different from a typical CTAB micelle (Chaimovich et al., 1985).

MECHANISMS OF REACTIONS OF 0 - L A C T A M ANTIBIOTICS

227

Catalysis by micelles of the hydroxide-ion catalysed hydrolysis of substrates appears to be qualitatively understood on the basis of a concentration effect of reactant on, or around, the micelle surface and need not necessarily involve a difference in the free energies of activation in the micelle and bulk phase. That is not to say that the cationic micelles could not and do not cause electrostatic stabilisation of the transition state. The cationic micelle surface can act as an electrostatic sink for the anionic intermediate leading to its stabilisation, but a rate enhancement requires preferential stabilisation of this intermediate compared with the reactant. The small rate enhancement of the micelle catalysed reaction, about 50-fold, is equally well explained by considering that the increased concentration of reactants at the micelle surface leads to a higher observed rate. Incorporation of the reactants into a limited volume decreases the entropy loss that is associated with bringing reactants together in the transition state and this leads to an increase in the pseudo first-order rate constants in the presence of surfactant micelles. Cationic micelles of CTAB have also been shown to facilitate the alkaline hydrolysis of the cephalosporin, cephalexin (Yatsuhara er ul., 1977). Added salts decrease the rate of the CTAB micelle catalysed alkaline hydrolysis of benzylpenicillin (Gensmantel and Page, 1982a). The salt effect can be considered to be due to competitive binding of the anions with the micelle. Increasing the unreactive anion concentration displaces hydroxide ion bound in the Stern layer leading to a reduction in the observed rate. If the assumption is made that the inhibition is competitive but only between the added anion and hydroxide ion, then the equation derived by Romsted can be applied to the kinetic data. The association constant, K,, between benzylpenicillin and micelle has been estimated to be 300 M - and the exchange constant, Ki,for bromide relative to hydroxide ion 25 M - ' . The relative degree of anion inhibition is that the larger the anion the lower its charge density and the larger the inhibition. The hydrolysis reaction is also inhibited by the addition of the hydrolysis product, benzylpenicilloate [30], a dianion which appears to bind no more tightly to the micelle than does benzylpenicillin itself. In benzylpenicilloate there are two carboxylate anions, yet the inhibition which results from increasing its concentration is similar to that caused by increasing the benzylpenicillin concentration. The effect of acetate ions is not large and it appears that carboxylate anions are rather ineffective anions in terms of binding with the CTAB micelle relative to simple inorganic anions. The attraction of organic molecules into a micelle can be due to both electrostatic and hydrophobic interactions. The rate of the hydroxide-ion catalysed hydrolysis of benzylpenicillin in the presence of micelles of CTAB is sensitive to electrolytes (Gensmantel and Page, 1982a), supporting the idea of electrostatic interactions between substrates and the micelle surface. The importance of other effects has been

228

MICHAEL I. PAGE

demonstrated by modifying the 6-P-side chain of penicillin to increase the substrate lipophilicity and hence the micelle-substrate hydrophobic interaction (Gensmantel and Page, 1982b). The second-order rate constants for the hydroxide-ion catalysed hydrolysis of 6-substituted penicillins are given in Table 9, and are independent of the alkyl substituent. Conversion of the 6-amino substituent to an acylamido group increases the rate constant ca 3-fold. The rate maximum in the rate-surfactant concentration profiles for the base catalysed hydrolysis of alkylpenicillins in the presence of CTAB moves to a lower surfactant concentration with increasing substrate lipophilicity, and the rate “maximum” is dependent on the penicillin concentration. Increasing the 6 4 acylamino chain length increases the lipophilic character of the substrate and increases the binding constant Ks and the rate enhancement (Gensmantel and Page, 1982b). Increasing the 6-acylamino chain length of the penicillin substrate not only decreases the surfactant concentration at which the maximum rate is observed but also results in a slightly increased maximal rate. The first effect may be rationalised on the basis of increasing affinity of the substrate for the micelle phase brought about by the increased hydrophobic interaction when the 6-acylamino side chain is increased in length. The second aspect, that of different rate maxima, must be more subtle. Increasing the surfactant concentration should eventually lead to all the substrate being associated with the micelle and, since the substrates hydrolyse in water with similar second-order rate constants, then the same rate maximum would be expected for each substrate, if, as generally accepted the rate constant within the micelle is similar to that in the aqueous phase. For compounds with lower affinities for the micelle, it is necessary to use higher concentrations of surfactant to incorporate all the penicillin substrate. Increasing the surfactant concentration also increases the concentration of unreactive counterion, and it is probably the displacement of reactants from the micelle surface by bromide ion that causes different rate maxima to be achieved for different substrates. Catalysis occurs below the cmc of CTAB and is most marked for the more lipophilic substrates, suggesting that induced micelle formation may be occurring (Gensmantel and Page, 1982b). The CTAB catalysed hydrolysis of penicillin derivatives appears to exhibit some degree of specificity. Increasing the hydrophobicity of the 6-P-side chain increases micellar catalysis. The association of the penicillin substrate with the micelle is presumably the result of interactions similar to those that give micelles stability relative to their monomeric form in aqueous solution; hence the not unexpected increase in substrate binding with increased lipophilicity of the molecule. It appears that once the 6-P-side has been

TABLE9 Summary of the data for the hydroxide-ion catalysed hydrolysis of penicillin derivatives in the presence and absence of micelles of cetyltrimethylammonium bromide at 30°C"

6P-Aminopcnicillanic acid 6-Methylpenicillin 6-Et h y lpenicillin 6-Propylpenicillin 6-Butylpcnicillin 6-Pen ty lpenicillin 6-Heptylpenicillin 6-Nonylpenicillin 6-Undecylpenicillin 6-Benzylpenicillin 6-Benzylpenicillin methyl ester

0.039 0.138 0.1 I6 0.109 0.1 I6 0.121 0.126 0.132 0.1 15 0.137 3.69

0.1 i 0.72 1.05 1.42 I .68 1.92 2.55 2.32 2.06 2.20 48.0

10 40 90 180 250 280 320 330 350 300 145

1.1

28.8 94.5 256 420 538 816 766 72 1 660 6660

2.8 5.2 9.1 13.0 14.5 15.9 20.2 17.6 17.9 16.1 13.0

28 208 819 2340 3625 4452 6464 5808 6265 4830 1885

" Gensmantel and Page, 1982b Second-order rate constant for hydrolysis in the absence of micelles, ionic strength I = 0.05 M Apparent sccond-ordcr ratc constant for hydrolysis in thc prcscncc of CTAB micelles, 0.05 M-NaOH and 2 x IW4 M penicillin Apparent binding constant of penicillin dcrivative to micellc

0

x2 -

,, N+, ,

-

..,

FIG. I2 Hypothetical orientation of 6-P-aminopenicillanic acid bound to micelles of CTAB (Gensmantel and Page. 1982b)

0

bC/

,o-

I

FIG. 13 Hypothetical orientation of alkylpenicillin bound to micelles of CTAB (Gensmantel and Page, 1982b)

MECHANISMS OF REACTIONS OF

0- LACTAM ANTIBIOTICS

231

extended to CH3(CH2),CONH-, further extension does not significantly increase the binding constant. It is interesting to note that there is no evidence of the longer chain compounds pulling the whole penicillin molecule into the interior of the micelle. The polar compound 6-P-aminopenicillanic acid is only weakly bound to the micelle, and electrostatic interactions may be all that exist between the substrate and micelle. Figure 13 illustrates schematically how increasing the length of the 6-P-acylamino side chain increases the hydrophobic interaction between the substrate and micelle and may thus alter the major orientation compared with that adopted by 6-APA as shown by comparison of Figs. 12 and 13. The p-lactam carbon is suitably situated for reaction with hydroxide ion and the carbonyl oxygen of the 6-amide linkage is positioned to allow some electrostatic interaction between the electron density on the carbonyl oxygen and the micelle surface. Figure 13 also explains why forming the methyl ester of benzylpenicillin leads to only a small reduction in the micelle-substrate binding constant relative to that for benzylpenicillin. Because the carboxylate anion points away from the micelle surface, it can have only a weak electrostatic interaction with the micelle surface. There have been relatively few studies of the micellar catalysed hydrolysis of amides and the effects are small (Gani and Yiout, 1978; Broxton and Duddy, 1979; Anoardi and Tonellato, 1977; O'Connor and Tan, 1980). From the mechanism of micellar catalysis outlined in Scheme 6 the ratio k,/k, gives the difference between the free energy of activation in the micellar phase and in the bulk aqueous phase. For bimolecular reactions an apparent rate enhancement of lo3 to lo4 can result from the higher concentration of reactants in the smaller volume of micelles given by RT In V,/ V,. where V , and V , are the respective volumes of micelle and aqueous phases. This acceleration can occur even if the rate constants within the two phases are identical. To observe this maximum rate enhancement resulting from a simple concentration effect, the free energy of transfer of the reactant from the aqueous to the micellar phase must be more than enough to offset the loss of entropy from its restriction to a smaller volume within the micelle. A comparison of the constant k,K, with k, is dependent upon the choice of standard state because the micelle catalysed reaction is a higher order process. For a bimolecular reaction within the micelle, k,K, has units of concentration- time- and represents the free energy difference between the reactants and micelle in the bulk aqueous phase and the transition state in the micellar phase. The analogous reaction in the absence of micelles proceeds with a rate constant k , with units of concentration-' time-'. The ratio k,K,/k, has units of concentration and represents the free energy of transfer of the transition state from the aqueous phase to the micellar phase. Relative values of k , give the relative free energies of binding the

232

MICHAEL I. PAGE

substituent to the micelle in the ground state and transition state; relative values of K , give the free energy of transfer of the substituent from the aqueous to the micellar phase in the ground state and relative values of k,K, give the free energy of transfer of the substituent from the aqueous phase in the ground state to the micellar phase in the transition state. The comparison of the micelle and non-micelle catalysed hydrolysis of penicillin derivatives is given in Table 9. The data refer to 0.05 M sodium hydroxide but the apparent rate enhancements would be greater at lower hydroxide-ion concentration (Gensmantel and Page, 1982a). The logarithm of the binding constants K, show a non-linear dependence upon the Hansch n-substituent constant for the 6-alkyl side chain. This non-linear relationship is reflected in an apparent decrease in the free energy of transfer of a methylene unit from water to the micelle with a maximum value of 0.48 kcal mol- for transfer in the ground state and a maximum of 0.71 kcal mol-’ for transfer in the transition state (Gensmantel and Page, 1982b). A saturation phenomenon with respect to increasing alkyl hydrophobicity of the substrate is not always observed and depends upon the structure of the rest of the substrate (Gensmantel and Page, 1982b). It has been estimated (Molyneux et al., 1965) that the free energy change for the complete transfer of a single methylene unit from water to the micellar phase is 0.65 kcal mol-’ which corresponds to a maximum rate or equilibrium difference of 3 at 25°C. The free energy of transfer of a methylene group from water to a non-polar liquid is about I .O kcal mol- (Nelson and DeLigny, 1968) and that to an enzyme from 2.1 to 3.8 kcal mol- (Page, 1976, 1977). The smaller value for transfer to micelles compared with enzymes presumably results from the “loose” interactions between the micelle, composed of several molecules of surfactant separated by their van der Waals radii, and the substrate compared with the “tight” interactions available from the substrate molecule and one molecule of enzyme, composed of many atoms closely packed together (Page, 1984b).

’

’

10 Cycloheptaamylose catalysed hydrolysis

Cycloamyloses (cyclic a- 1,4-linked oligomers of D-glucose) have a toroidal or “doughnut”-shaped structure. The primary hydroxy groups are located on one side of the torus while the secondary ones lie on the other side. Relative to water the interior of the cycloamylose torus is apolar. The catalytic properties of cycloamyloses depend on the formation of inclusion complexes with the substrate and subsequent catalysis by either the hydroxy, or other groups, located around the circumference of the cavity (Komiyama and Bender, 1984; Page and Crombie, 1984).

MECHANISMS OF REACTIONS OF 8 - L A C T A M ANTIBIOTICS

233

Under mildly alkaline conditions and in the presence of excess cycloheptaamylose the rate of degradation of penicillin is increased 20-90-fold compared with the rate of alkaline hydrolysis (Tutt and Schwartz, 1971). Michaelis-Menten kinetics are observed which are indicative of complex formation. The apparent binding constant of 6-substituted penicillins varies little with the length of the alkyl side chain although it is increased about 1 0-fold for diphenylmethyl penicillin. The reaction is catalytic and hydrolysis proceeds by the formation of a penicilloyl-p-cyclodextrin covalent intermediate, i.e. ester formation, by nucleophilic attack of a carbohydrate hydroxyl on the p-lactam.

11

The aminolysis of p-lactam antibiotics

The reaction of amines with penicillins to give penicilloyl amides (Scheme 7) is of interest because the major antigenic determinant of penicillin allergy is the penicilloyl group bound by an amide linkage to E-amino-groups of lysine residues in proteins (Levine and Ovary, 1961; DeWeck and Bulm, 1965; Parker et al., 1962). The formation of the penicilloyl haptenic groups could conceivably occur by the direct aminolysis of penicillin (Schneider and DeWeck, 1966; Batchelor et al., 1965) or by the aminolysis of penicillenic acid formed from a rearrangement of penicillin (Levine, 1961 ; Bundgaard, 1980; De Weck, 1962; Schwartz, 1969) or by the reaction of amines with the ketene [47] (Gensmantel et al., 1978) formed by an elimination mechanism (Scheme 7). The aminolysis of penicillin is also worthy of study because the reaction is an amide exchange, a normally difficult process but one which occurs readily with p-lactams (Blackburn and Plackett, 1973). Carbon-nitrogen bond fission in amides usually requires protonation of the nitrogen to avoid expulsion of the unstable anion, but in p-lactams this process is accompanied by a large release of strain energy which modifies the requirements for catalysis compared with normal amides. Another important difference between carbon-nitrogen bond fission in p-lactams compared with that in amides is that the latter may be accompanied by a more favourable entropy change as the molecule fragments into two separate entities (Page, 1973). Because of the rigidity and shape of the penicillin molecule it is a suitable substrate to study the effectiveness of intramolecular catalysis and, in particular, to elucidate any preferred direction of nucleophilic attack upon the p-lactam carbonyl group (Martin et al., 1978). There have been several studies on the self aminolysis of penicillins containing amino groups which leads to dimerisation and polymerisation products (Bundgaard, 1977a,b; Larsen and Bundgaard, 1977, 1978a,b). The

MICHAEL I. PAGE

234

protein-NH,

CO,H

_ 1u^ protein

[I1

prokin - N H 2

Scheme 7

kinetics of the aminolysis of penicillins with protein has also been reported (Bundgaard and Buur, 1983). INTERMOLECULAR GENERAL BASE CATALYSIS

The aminolysis of penicillin is a substitution reaction in which an acyl group is transferred from one amino group to another. This reaction requires at least two proton transfers, proton removal from the attacking amine and proton addition to the leaving amino group. These proton transfers are facilitated by buffers (Morris and Page, 1980a), and the kinetic importance of such catalysis is usually related to the observation that “catalysis occurs where it is most needed”. Buffer catalysis is needed in the aminolysis of penicillin because covalent bond formation and fission between heavy atoms is accompanied by large changes in the acidity and basicity of the reacting groups (Jencks, 1976). If a proton is not removed from the attacking amine

235


at some stage during the reaction the acidity of the NH group would change by ca 40 pK units, from ca 30 in the reactant to ca - 10 in the hypothetical N-protonated amide [48]. Similarly, fission of the p-lactam C-N bond causes a change of ca 35 pK units in the basicity of the leaving amino group.

These large changes in pK can give rise to unstable intermediates such as [49] and [50], and buffer catalysis is observed because it can increase the rate of the reaction by either trapping such unstable intermediates or by stabilising or bypassing the transition states leading to their formation (Jencks, 1976). RCONH

RCONH

+fly co;

RH,N

[491

- 0 - 3

RGH

co; [501

It is generally accepted that acyl-transfer reactions involve the intermediate formation of tetrahedral addition compounds, such as [50],i.e. bond formation to the attacking group occurs before bond fission to the leaving group. The bond-breaking process may occur after, during, or even before (Page and Jencks, 1972a) the rate-limiting step. Further complications in the elucidation of the detailed mechanism of acyl-transfer reactions arise from the problem of the timing of the proton-transfer steps, i.e. are they concerted with, or separate processes from, covalent bond changes between heavy atoms? The aminolysis of penicillins and cephalosporins is a stepwise process catalysed predominantly by bases which remove a proton from the attacking amine. The evidence for the reversible formation of a tetrahedral intermediate is kinetic and based on linear free-energy relationships (Page, 1984a). The aminolysis of benzylpenicillin at 30°C in aqueous solutions of the amine follows the rate law (8), where kobsis the observed pseudo first-order rate constant for the disappearance of penicillin and k, is the second-order Rate = kabs = kJOH-1 [Pen1

~

+ k,[RNH,] + kb[RNH,I2 + k,,[RNH,][OH-]

(8)

236

MICHAEL I PAGE

rate constant for the hydrolysis reaction (Tsuji et al., 1975; Bundgaard, 1976d; Morris and Page, 1980a). The genera! acid-catalysed aminolysis of penicillin makes a negligible contribution to the observed rate. The dominant form of buffer catalysis in the aminolysis is general base catalysis. The relative importance of the terms in (8) depends on the basicity and the concentration of the amine and the pH. For strongly basic amines the amine catalysed (k,) and the hydroxide-ion catalysed (koH) terms contribute most to the observed rate with the k , term, of course, being more important with increasing concentration of amine. Consequently, the rate constants k, for the uncatalysed reactions of basic monoamines are not of high precision. For the more weakly basic amines aminolysis occurs mainly through the uncatalysed (k,) and amine catalysed (k,) pathways because of the low concentration of hydroxide ion. The hydroxide-ion catalysed term (koH) makes a negligible contribution to the observed rate of aminolysis in buffers of amines with pKa < ca 9, and can only be determined in solutions of sodium hydroxide (Morris and Page, 1980a). Linear-free energy relationships have been determined by varying independently the reactivity of the amine nucleophile and the catalyst. A plot of k, for the general base catalysed aminolysis of benzylpenicillin for a series of primary monoamines against the pKa values of the amines gives a straight line, the slope of which, the Brernsted p-value, is 1.09 f 0.09 (Morris and Page, 1980a). This means that the reaction behaves as if, in the transition state, approximately unit positive charge is developed and is distributed between the nucleophilic and the catalysing amine molecules. A Brernsted p-value of 0.68 at 60°C has been reported for this reaction, but this was based on a series of diverse amines including imidazole (Tsuji et al., 1975). Another reported value of 0.82 at 35°C was derived from results that included data for glycine ethyl ester which is known to hydrolyse under the reaction conditions (Bundgaard, 1976d). The Brernsted p-value of ca unity is indicative of a transition state in which full covalent bond formation has taken place between the nitrogen of the attacking amine and the carbonyl carbon and which carries the positive charge on either the nitrogen of the nucleophilic amine or on the catalytic amine molecule. The simplest mechanism that is consistent with this observation is shown in Scheme 8. The first step involves nucleophilic attack of the amine to form the tetrahedral intermediate T i , for which there is independent kinetic evidence (see later). However, the intermediate T * breaks down rapidly to starting materials by expulsion of the attacking amine ( k - Catalysis of the reaction occurs by the formation of an encounter complex between T f and the basic catalyst B (Scheme 8). For general base catalysis by amines, B is an amine. Subsequent proton transfer from T i to B forms T- which then breaks down to products.

MECHANISMS OF REACTIONS OF

B- LACTAM

ANTI B lOTl CS

237

HNR

TScheme 8

The Brmsted p,,,-value for the hydroxide-ion catalysed aminolysis of benzylpenicillin (koH) for a series of primary monoamines is 0.96 (Morris and Page, 1980a). This value also indicates that the reaction behaves as if a unit positive charge is developed on the attacking amine in the transition state. The assignment of charge density is unambiguous, unlike the case for the general base catalysed reaction, and the simplest interpretation of the p,,,-value is that the attacking amine resembles its conjugate acid, i.e. is fully protonated, in the transition state. This is compatible with the mechanism of Scheme 8 in which the bond between the attacking amine and the carbonyl carbon is fully formed. The P,,,-value indicates the location of the proton in RCONH

-0kC3 RNH

kOi

[511

the transition state, for there can be little or no proton transfer from the attacking amine to the hydroxide-ion catalyst. It is consistent with ratelimiting diffusion-controlled encounter of the tetrahedral intermediate T * and hydroxide ion giving [51]. Hydrazine shows an enhanced nucleophilic reactivity towards penicillin compared with amines of similar basicity which is attributed to the a-effect. This has allowed a study of the effect of varying the basicity of the catalyst with a constant nucleophile even in the presence of strongly basic catalysts.

238

MICHAEL I. PAGE

For example, catalysis of the reaction of hydrazine with benzylpenicillin occurs even with the strongly basic amine propylamine (Morris and Page, 1980b). There is a non-linear dependence of the rate of hydrazinolysis of benzylpenicillin upon the basicity of both oxygen and nitrogen base catalysts. For strongly basic catalysts there is little dependence of the rate constants upon basicity and the Brransted p-value is < 0.2. However, catalysis by weak bases shows a much stronger dependence upon the basicity of the catalyst with p 2 0.8. A curved or non-linear Brransted plot is required to describe the behaviour of both oxygen and nitrogen bases. The large sensitivity of the rate constants to base strength for weakly basic catalysts indicates that the catalyst resembles its conjugate acid in the transition state, i.e. there is a large amount of, or complete, proton transfer to the catalyst in the transition state. For strongly basic catalysts the small sensitivity of the rate constants upon base strength suggests that the catalyst resembles its free unprotonated basic form in the transition state. The simplest mechanism compatible with the observation involves the formation of an unstable dipolar tetrahedral addition intermediate T i which rapidly reverts to the starting materials by expulsion of the attacking Reaction only proceeds if the intermediate is trapped by an amine, (k1). encounter with a base that results in proton transfer to form the anionic intermediate T - , which rapidly breaks down to products (Scheme 8). Proton transfer between electronegative atoms is thought to occur by a stepwise process involving the diffusion-controlled encounter of the proton donor and acceptor, followed by proton transfer itself and then diffusion apart (Eigen, 1964). Proton transfer itself ( k 3 ) is not usually rate-limiting. The application of these suggestions to the mechanism of aminolysis of penicillin provides an explanation for the non-linear Brernsted plot. When the tetrahedral intermediate, T i , is a stronger acid than the conjugate acid of the basic catalyst, proton transfer is thermodynamically favourable. The ratelimiting step will therefore be the diffusion-controlled encounter of T i and the catalyst ( k , ) and the observed rate will be independent of the basicity of the catalyst (Jencks, 1976). However, for weakly basic catalysts proton transfer is thermodynamically unfavourable and the rate-limiting step changes to the diffusion apart of the deprotonated intermediate, T - , and the protonated catalyst ( k J . This kinetic scheme explains the large dependence of the rate upon the basicity of the catalyst for weakly basic catalysts and its insensitivity for strongly basic catalysts (Morris and Page, 1980b). In addition to this change in rate-limiting step deduced from non-linear free energy relationships by changing the basicity of the catalyst, another change has been observed directly from the kinetics of the hydroxide-ion catalysed aminolysis of benzylpenicillin (Gensmantel and Page, I979a). In aqueous sodium hydroxide the aminolysis occurs largely by the k,, pathway


239

(8). There is a non-linear dependence of the apparent second-order rate constants upon the concentration of hydroxide ion. At low concentrations of hydroxide ion the rate is first-order in hydroxide ion and the initial slopes give values of k,, which agree well with those determined at lower pH in buffer solutions (Gensmantel and Page, 1979a; Morris and Page, 1980a). At high concentrations of hydroxide ion the rate becomes independent of the concentration of hydroxide ion. This change in the kinetic dependence on hydroxide ion is indicative of a change in the rate-limiting step of the reaction which, in turn, requires that there be at least two sequential steps in the reaction. One of these steps is rate-limiting at low concentrations of hydroxide ion and the transition state for this step contains hydroxide ion, or its kinetic equivalent. The other step is rate-limiting at high concentrations of hydroxide ion but the transition state for this step does not contain hydroxide ion. The existence of two sequential steps demands that there be an intermediate in the reaction which is probably the tetrahedral intermediate T'. The mechanism of Scheme 8 is compatible with this observation. At low concentrations of hydroxide ion the rate of collapse of the tetrahedral intermediate to reactants must be faster than its reaction with hydroxide ion ( k 1 9 k,[OH-I); the observed rate constant is dependent upon the concentration of hydroxide ion with k,, the diffusion-controlled step, being rate-limiting. The calculated pK,-values for the protonated amine of the tetrahedral intermediates are well below that for water. Proton transfer from the tetrahedral intermediate to hydroxide ion is therefore in the thermodynamically favourable direction and it is to be expected that the rate-limiting step for this process is the diffusion-controlled encounter of the proton donor and acceptor. At high concentrations of hydroxide ion the tetrahedral intermediate and hydroxide ion diffuse together faster than the intermediate collapses back to reactants (k,[OH-]9 k Under these conditions the observed rate constant is independent of hydroxide ion concentration and k , , the rate of formation of the tetrahedral intermediate, is rate-limiting (Gensmantel and Page, 1979a). Values of k , thus determined for a series of amines yield a Brernsted p,,, of 0.3. This indicates that the reaction behaves as if there is a development of a charge of ca + 0.3 on the attacking amine nitrogen in the transition state [52], which must therefore occur early along the reaction coordinate with little C-N bond formation. Assuming that the diffusion controlled step k , has a value of 10" M - l s - l , values of k - , and the equilibrium constants for the formation of the tetrahedral intermediates have been obtained (Gensmantel and Page, 1979a). The rates of expulsion of the attacking amine from the tetrahedral intermediate to regenerate the reactants ( k - , ) are very rapid, ca

MICHAEL I. PAGE

240

+

n

0.3 RNH,---$-NII 0 0.3 -

[521

109-10'0s-'. Although these rate constants are very large they are of the order of magnitude that have been postulated for the breakdown of tetrahedral intermediates formed in acyl transfer reactions. Similarly, the equilibrium constants for the formation of the tetrahedral intermediates, T i , have been obtained. These vary substantially with the basicity of the amine from 4 x lo-'' M - ' for 2-cyanoethylamine to 9 x 1 0 - 9 ~ - 1for propylamine. The Brsnsted p,,,-value for the equilibrium is 0.9. This provides experimental support for the Brsnsted p-value of 1.0 that is often postulated for the formation of the tetrahedral intermediate from amines and carbonyl groups in which the amine nitrogen develops a unit positive charge, and presumably resembles the conjugate acid of the amine in structure and in its stability dependence upon substituents. Similar observations have been made in the aminolysis of cephalosporins (Page and Proctor, 1984), and these are discussed in Section 12. The aminolysis of cephalosporins follows a similar mechanism to that for penicillins (Proctor and Page, 1984). Although it has been suggested that the hydroxide-ion catalysed aminolysis involves proton transfer concerted with nucleophilic attack (Bundgaard, 1975), the Brsnsted p,,,-values of ca 1.O are consistent with the stepwise mechanism. This is also supported'by the non-linear dependence of the rate of aminolysis of cephalosporins upon hydroxide ion concentration (Proctor and Page, 1984) (see Section 12). The partitioning of the tetrahedral intermediate (T', Scheme 8) formed by nucleophilic attack upon a p-lactam is controlled by the ease of exocyclic versus endocyclic bond fission. Endocyclic C-N bond fission is favoured by the release of the strain energy of the four-membered ring. Expulsion of the attacking nucleophile by exocyclic bond cleavage is accompanied by a relatively favourable entropy change because two molecules are generated from one (Page, 1973). Except in the presence of strongly acidic catalysts, general acid catalysed breakdown of the tetrahedral intermediate to products by proton transfer to the p-lactam nitrogen is a relatively unimportant pathway in aminolysis. This is expected because of the weakly basic p-lactam nitrogen in the tetrahedral intermediate (Morris and Page, 1980a). That expulsion of the attacking amine nucleophile from [50] occurs more readily than fission of the p-lactam C-N bond is confirmed by the observation that 2-azetidinylideneammonium salts [53] react with hydroxide to give 2-azetidinones [54] presumably through the intermediate formation


241

of [55] (Poortere et a/., 1974; Agathocleous et al., 1985). Similarly, several synthetic reactions have been reported in which the exocyclic p-lactam oxygen is exchanged without fission of the p-lactam ring (Gilpin et a/., 1981; Wojtkowski et a/., 1975). The rate law for hydrolysis of [53] shows a third order term which is first order in substrate, hydroxide and carbonate ions and indicates that the rate-limiting step must involve breakdown of the tetrahedral intermediate [55] and that formation of [55] must be reversible. The simplest explanation is that C-N cleavage occurs by general acid (HCO J) catalysed breakdown of the anion of the tetrahedral intermediate, [ 5 5 ] (Page ef a/., 1987).

+pNzr+?

-N

I

OH

endocyclic

[541 INTRAMOLECULAR GENERAL BASE CATALYSIS

Intermolecular general base catalysis in the aminolysis of penicillins is a major pathway for product formation (p. 234). It is not surprising therefore that intramolecular general base catalysed aminolysis has been observed (Schwartz, 1968). The rate constant k, for the reaction of 1,2-diaminoethane with benzyl.penicillin is ca 30-fold greater than that predicted for a monoamine of the same basicity from the Brernsted plot (Martin el a/., 1976; Morris and Page, 1980a). The rate enhancement is interpreted as evidence for intramolecular general base catalysis of aminolysis by the second amino group in 1,2diaminoethane [56]. Proton transfer occurs from the amino group that acts as the nucleophile to the terminal amino group acting as a general base.

[561

Most of this rate enhancement is a result of the greater basicity of the amino group compared with water. Very little of the rate enhancement is attributable to intramolecularity, with the catalyst being covalently linked to the nucleophile. This is evident from the effective molarity of ca 1 moll- for the reaction which is obtained by dividing the second-order rate constant, k,, for the reaction of 1,2-diaminoethane with penicillin by the third-order rate constant, k,, for intermolecular catalysis of aminolysis by a second molecule of amine of similar basicity. The effective molarity is the concentration of catalysing amine required to give the same rate of reaction as the diamine. Similar, small effective molarities have been observed for intra-

242

MICHAEL I. PAGE

molecular general acid base catalysed reactions (Page, 1973; Kirby, 1980). Intramolecular catalysis is observed because of the importance of general base catalysis in these reactions compared with uncatalysed aminolysis. By analogy with the intermolecular general base catalysed reaction, the rate-limiting step in the intramolecular reaction is probably a conformational change. It appears that the dominant contribution to the low effective molarity is the “loose” transition state of the intermolecular reaction. The rate-limiting step of the intermolecular general base catalysed aminolysis of penicillin is the diffusion-controlled encounter of the tetrahedral intermediate, T i , with the base (Scheme 8). The transition state is thus very “loose” and the bimolecular step, k, in Scheme 8, will be associated with a small entropy change giving rise to the low erfective concentration of the intramolecular reaction (Page, 1977; Morris and Page, 1980a). A suitably placed amino group within the penicillin molecule, as opposed to one in the attacking amine nucleophile, could conceivably also act as an intramolecular general base catalyst for aminolysis. In fact, the aminolysis of 6-P-aminopenicillanic acid [ 101 occurs predominantly by an uncatalysed pathway. The term k , in rate law (8), which is second order in amine and predominant for the aminolysis of penicillin, is of minor significance for the reaction of monoamines with 6-P-aminopenicillanic acid. This has been interpreted as evidence for intramolecular general base catalysis [57] (Schwartz and Wu, 1966). 7..

H

NH,

a3

0 ’

t0,H

COY [571

[581

However, the second-order rate constants, k,, for the reaction of monoamines with benzylpenicillin and 6-P-aminopenicillanic acid are similar; there is no rate enhancement. Furthermore, the aminolysis of penicillanic acid [58] also does not show a significant general base catalysed term in the rate law (Gensmantel and Page, 1979b). Clearly, this predominance of the uncatalysed aminolysis pathway cannot be attributed to intramolecular general base catalysis. The aminolysis of 6-P-aminopenicillanic acid [ 101 in solutions of sodium hydroxide has enabled the rate constants k, and k-, and the equilibrium constant K (Scheme 8) to be deduced. The values of k, for the formation of the tetrahedral intermediates from 6-P-aminopenicillanic acid and from

243


benzylpenicillin are similar, which is again indicative of the lack of intramolecular general base catalysis by the neighbouring amino group in [ 101 (Gensmantel and Page, 1979a). The reason for the absence of significant intramolecular general base catalysis in the aminolysis of 6-P-APA and penicillanic acid is discussed on p. 244. INTRAMOLECULAR GENERAL ACID CATALYSIS A N D THE DIRECTION

OF N U C L E O P H I L I C A T T A C K

The rate constant k, for the reaction of penicillin with the monocation of 1,2-diaminoethane is cu 100-fold greater than that predicted from the B r ~ n s t e dplot for a monoamine of the same basicity. The rate enhancement is attributed to intramolecular general acid catalysis of aminolysis by the protonated amine (Morris and Page, 1980a; Martin et ul., 1979). Breakdown of the tetrahedral intermediate, Ti,is facilitated by proton donation from the terminal protonated amino group to the p-lactam nitrogen [59]. It is not known whether proton transfer and carbon-nitrogen bond fission are concerted or occur by a stepwise process. RCONH

NHCOR

According to the theory of stereoelectric control of Deslongchamps (1975), the breakdown of tetrahedral intermediates is facilitated by the lone pairs of the heteroatoms attached to the incipient carbonyl carbon being antiperiplanar to the leaving group. Application of this theory to the microscopic reverse steps predicts that the direction of nucleophilic attack on the carbonyl carbon be such that the lone pairs on the heteroatoms will be antiperiplanar to the attacking group. Penicillins have a fairly rigid structure because of the fusion of the p-lactam and the thiazolidine rings giving a V-shaped molecule. A consequence of the non-planarity of the fused bicyclic ring system is that the electron density of the lone pair of the S-lactam nitrogen will be concentrated heavily on the a-face of the penicillin molecule [ 171 and particularly of the tetrahedral intermediate [60]. According to the theory of stereoelectronic

MICHAEL I. PAGE

244

control nucleophilic attack on penicillins should, therefore, take place from the P-side. However, this face is sterically hindered and it has been suggested that nucleophilic attack may therefore take place from the less hindered a-side (Martin et al., 1979; Gensmantel and Page, 1979a). The observation of intramolecular general acid catalysis in the reaction with the monocation of 1,2-diaminoethane gives an indication of the direction of nucleophilic attack upon penicillin. In order that ready proton transfer takes place from the protonated amine to the p-lactam nitrogen, it is essential that the tetrahedral intermediate has the geometry shown [60]. Although intramolecular general acid catalysis could conceivably take place if the amine attacked from the p-face [61], this would involve considerable non-bonded interactions and/or the proton transfer taking place through one or more water molecules. Further evidence for nucleophilic attack taking place from the a-face comes from the absence of intramolecular NHCOR

general base catalysis in the aminolysis of 6-P-aminopenicillanic acid (p. 242). That the lone pair on the p-lactam nitrogen takes up the geometry with respect to the carboxy-group shown in [60] is supported by the observation that copper(I1) ions catalyse the aminolysis of penicillin by coordination to the p-lactam nitrogen and the carboxy-group, thus stabilising the tetrahedral intermediate (Gensmantel et al., 1978). The low effective molarity of intramolecular aminolysis from the p-side is also consistent with this side being sterically unfavourable (p. 249). Thus nucleophilic attack on penicillins, at least by amines, appears to take place from the least hindered a-side in disagreement with the prediction of the theory of stereoelectronic control. It is unlikely that a-attack would give the stereoisomer predicted by stereoelectronic control because this would introduce a highly strained transfused bicyclic system. UNCATALYSED AMINOLYSIS

The Br~nstedp,,,-value for the uncatalysed aminolysis of penicillins is 1.O (Morris and Page, 1980a), which indicates that the reaction behaves as if a unit positive charge is developed on nitrogen in the transition state. The

MECHANISMS OF REACTIONS OF b - L A C T A M ANTIBIOTICS

245

uncatalysed pathway characterised by k , could represent either a purely uncatalysed reaction of amine and penicillin or solvent catalysis with water acting either as a general base, removing a proton from the attacking amine, or as a general acid, donating a proton to the (j-lactam nitrogen. The k , pathway cannot represent rate-limiting formation of the tetrahedral intermediate because the (j,,,-value for this is known to be 0.3, and the rate constants k , (Scheme 8) are known to be much greater than the observed k,values (Gensmantel and Page, 1979a). The rate constant k , for hydrazine divided by the concentration of water, 55M, gives a third-order rate constant with a large positive deviation from the Brernsted plot for general base catalysed hydrazinolysis (Morris and Page, 1980b); this indicates that water is not acting as a proton acceptor. It is not easy to distinguish between uncatalysed rate-limiting breakdown of the tetrahedral intermediate, T i , [62] and breakdown of the same intermediate general acid catalysed by water ~31.

' ' g c \ RNH,

O'PNRyH,LH

i

0

0 '

I

H

[621 [631

The equilibrium constant K , for the formation of T' from penicillin and propylamine is known to be 8.86 x 10-91mol-' (Gensmantel and Page, 1979a). Because the rate constant for the uncatalysed reaction, k,, is given by k , K , (Scheme 8), k , , the rate constant for the uncatalysed or water-catalysed breakdown of the tetrahedral intermediate, is k , / K , = 1.5 x 10's-l. The rate of protonation of the (j-lactam nitrogen of T' by water may be estimated from its pK,-value of 5.2 to be ca 103s-'. This means that the uncatalysed breakdown of T' cannot proceed by stepwise proton transfer from water to the (j-lactam nitrogen. However, it could occur either by a concerted mechanism - proton transfer from water occurring synchronously with carbon-nitrogen bond fission - or uncatalysed expulsion of the nitrogen anion. The rate of expulsion of the imidazolyl anion from the tetrahedral intermediate formed in the aminolysis of acetylimidazole is 2 106s-' (Page and Jencks, 1972b). If the strain energy of the p-lactam ring of ca 120 kJ mol - is relieved upon ring opening, then expulsion of the nitrogen as the anion should be treated as a leaving group of pK,-value of ca 10 rather than the normal value of 30 for ordinary amines. It is conceivable therefore that carbon-nitrogen bond fission occurs without protonation of the p-lactam nitrogen [62]. The uncatalysed aminolysis of cephalosporins is discussed in Section 12.

MICHAEL I. PAGE

246

The reason for the absence of'a significant general base catalysed term in the rate law for the aminolysis of 6-P-aminopenicillanic acid [lo] and penicillanic acid [58] is that the rate constants for the uncatalysed breakdown of the tetrahedral intermediates, k, in Scheme 8, are 13- and 50-fold, respectively, greater than that for benzylpenicillin (Gensmantel and Page, 1979b). The observed contributions of general base catalysed and uncatalysed aminolysis to the rate depend on the ratio of k, to k, [RNH,] (Scheme 8). For [lo] and [58] k, [RNH,] does not become greater than k, until the amine concentration is greater than 0.5 M. METAL-ION CATALYSED AMINOLYSIS

In aqueous solution in the presence of copper(I1) ion penicillin reacts with amines to form the corresponding penicilloyl amide. The kinetics of this reaction show a saturation phenomenon with the concentration of metal ion but are complicated by the complexation of copper(I1) ions with the amine (Gensmantel et al., 1978). A kinetic scheme compatible with the observed data is given in Scheme 9 and where B represents the amine, M copper(I1) ion, P penicillin, MP is the penicillin-copper(I1) complex and M B is the amineecopper(I1) complex. M

MB

+P

K,

MP

k,[B)

aminolysis products

hydrolysis products Scheme 9

The enormous rate enhancement brought about by copper(I1) ion is appreciated when aminolysis of penicillin occurs, for example, with propylamine at p H 4 when the concentration of free propylamine is only ca 10-7-10-8M. The rate enhancements of amines reacting with the penicillincopper(I1) complex compared with their reaction with penicillin alone are ca 4 x lo6 and lo7 respectively. These rate enhancements are similar to that for hydroxide ion, ca 9 x lo7, described earlier (Section 9). These large rate enhancements are attributable to the copper(I1) ion complexing with penicillin [26] because methylation of the free carboxylate group in benzylpenicillin reduces the rate of the copper(I1) ion-trifluoroethylamine reaction by ca lo3. There is no kinetic dependence upon the concentration of trifluoroethylamine which indicates that aminolysis does


247

not occur in this reaction. This observation may be rationalised by metal-ion coordination to the carboxy-group in penicillin which is reduced or does not occur in the methyl ester. However, the rate of reaction of the methyl ester is increased by ca lo3 in the presence of 2 x 10-3M copper(I1) ions and the rate is apparently first order in metal ion. This may be due either to weak coordination of the metal ion to the p-lactam nitrogen and the methoxycarbonyl group compared with penicillin itself, or to binding at another site in the penicillin molecule. The product of the reaction of the methyl ester in the presence of copper(I1) ions is not known (Gensmantel er af., 1978). The dependence of the rate constants for the attack of amine on the metal ion-penicillin complex upon the basicity of the attacking amine gives a Brernsted p-value of 0.87. This is taken to indicate that there is approximately a unit positive charge on the amine nitrogen in the transition state. A mechanism consistent with this involves the rate-limiting breakdown of the tetrahedral intermediate [64].

In the rate law for the metal-ion catalysed reaction there is no evidence of a term second order in amine which, if present, would indicate general base catalysis by a second molecule of amine. The coordination of the metal ion to penicillin apparently makes this normally dominant mode of catalysis unnecessary. If the mechanisms of the reactions of penicillin involves the intermediate formation of a keten [47] or penicillenic acid [38], then the products of the reaction should show deuterium incorporation at C(6) if the reactions are carried out in D,O. The nmr spectra of benzylpenicilloic acid and penicilloyl amides obtained from the hydrolysis and aminolysis of benzylpenicillin in D,O in the presence of copper(I1) ion shows no incorporation of deuterium at C(6). This indicates that the elimination-addition mechanism is not a major pathway for either of these reactions (Gensmantel er al., 1978). The presence of copper(I1) ion has little effect upon the rate of formation of penicillenic acid from benzylpenicillin. Furthermore, the rates of hydrolysis and aminolysis of benzylpenicillenic acid are retarded in the presence of copper(I1) ion. Since the observed rates of hydrolysis and aminolysis of penicillin in the presence of copper(I1) ion are at least lo3 times faster than the rate of formation of penicillenic acid from penicillin these reactions cannot

248

MICHAEL I. PAGE

occur through the intermediate formation of penicillenic acid (Gensmantel et ul., 1978). Zinc(I1) and tris-buffers are effective catalysts for the aminolysis of benzylpenicillin. It is suggested that this is due to formation of a ternary complex in which the metal ion binds both penicillin and tris. Nucleophilic attack of the ionised hydroxyl on bound tris forms a penicilloyl ester which may then react with tris to form a penicilloyl amide (Schwartz, 1982; Tomida and Schwartz, 1983). A kinetically equivalent mechanism, however, would simply involve nucleophilic attack of tris on the zinc-penicillin complex. IMIDAZOLE CATALYSED ISOMERISATION OF P E N I C I L L I N S

Unlike other amines, imidazole catalyses the isomerisation of benzylpenicillin to benzylpenicillenic acid [38] (Bundgaard, 1971a,b, 1972, 1976a). The rate law shows that the nucleophilic reaction with imidazole to give the intermediate penicilloylimidazole [65] is general base catalysed by another molecule of imidazole and general acid catalysed by the cqnjugate acid of imidazole.

-i:xcoi Q

RCONH

\

o’/c\N

[651

The mechanisms originally proposed for the formation of penicilloyl imidazole involved a concerted reaction in which nucleophilic attack by imidazole occurred simultaneously with proton transfer (Bundgaard, 1972; Yamana et al., 1975). More recently it has been suggested (Butler et al., 1982) that the mechanism is similar to the stepwise process proposed for the aminolysis of penicillins (Morris and Page, 1980a). Imidazole also catalyses the aminolysis of penicillins (Yamana et al., 1975; Bundgaard, 1976a). The formation of penicilloyl amides could occur by acyl transfer from the intermediate penicilloyl imidazole [65] to the amine or by aminolysis of penicillenic acid. The aminolysis of acylimidazoles is well known (Page and Jencks, 1972b), but it is claimed that intramolecular attack by the 6-side chain amido group to displace imidazole and to give the oxazolinone-thiazolidine and then penicillenic acid will be faster than intermolecular attack of amine (Yamana et al., 1975). Imidazole catalysed aminolysis of penicillin was therefore suggested to occur exclusively through the reaction of the amine with penicillenic acid. Oxazolinone formation from


249

N-benzoylglycinate esters occurs when there is a good leaving group as in phenyl esters (Williams, 1975). Acetylimidazole is more reactive than phenyl acetate (Oakenfull and Jencks, 1971) and so displacement of the imidazole by the intramolecular amido group is expected. However, penicillins incapable of forming penicillenic acid undergo an imidazole catalysed aminolysis, presumably via the intermediate formation of an N-penicilloylimidazole. Furthermore, penicillenic acid reacts with imidazole to form N-penicilloylimidazole suggesting that the latter may be the acylating agent for aminolysis (Bundgaard, 1976a). INTRAMOLECULAR AMINOLYSIS

There have been many reports of a suitably placed intramolecular amino group attacking the p-lactam of cephalosporins (Tsuji et al., 1981, 1983; Bundgaard, 1977b; Dinner 1977). Cephalosporins which have an a-amino group in the 7-amido side chain, e.g. cephalexin, form piperazine-2,5-diones by such a pathway [66], whilst analogous penicillins do not (Indelicato et al., 0

COI [661

1974). Ring closure is predominant at neutral pH and is subject to hydroxide ion and general acid-base catalysis. There is a non-linear dependence of the rate upon buffer concentration which is indicative of a change in rate-limiting step and evidence for the formation of an intermediate. At 35°C the rate constant for the uncatalysed intramolecular aminolysis of cephaloglycin is 6.3 x 10-5s-' and the second-order rate constant for the hydroxide ion catalysed reaction is 0.22 M-'s-' (Bundgaard, 1976b). The estimated rate constants for the equivalent intermolecular reactions for an amine of pK, = 7 are 2 x 1 0 - 6 M - 1 ~ - 1and 6 x 10-3M-2s-' respectively (Page and Proctor, 1984). This gives an effective molarity of only about 35 M for the intramolecular amino group which is very low compared with other intramolecular aminolysis reactions (Kirby, 1980; Page, 1973). Such a low value is probably due to the introduction of unfavourable steric strain in the intramolecular reaction and problems with keeping the side chain amido group coplanar. Nucleophilic attack on the (j-lactam of penicillins takes place from the a-side (p. 243) and intramolecular attack from the sterically

2 50

MICHAEL I. PAGE

hindered P-side is unfavourable, particularly because of interference from the 2-P-methyl group. It is not surprising therefore that intramolecular aminolysis by amino groups in the 6-P-side chain of penicillins does not occur (Indelicato et al., 1974) and that intermolecular aminolysis of cephalosporins is competitive with intramolecular aminolysis (Bundgaard, 1976~). 6-(N-Phenylureido)penicillanic acids undergo a rapid cyclisation to the isomeric 3-phenylhydantoin-thiazolidines by intramolecular nucleophilic attack of the ureido nitrogen anion on the p-lactam carbonyl carbon. The second-order rate constant for the hydroxide ion catalysed reaction is over 103-fold greater than that for hydroxide ion catalysed hydrolysis (Bundgaard, 1973). The rate enhancement is consistent with neighbouring group participation. 12 The stepwise mechanism for expulsion of C(3)-leaving groups in cephalosporins

That expulsion of the leaving group at C(3‘) is not generally concerted with C-N bond fission of the p-lactam when cephalosporins react with nucleophiles was described in Section 4. Despite experimental evidence (HamiltonMiller et al., 1970b; Page and Proctor, 1984; Agathocleous er al., 1985; Grabowski et al., 1985; Faraci and Pratt, 1984, 1985) that this is the case for both enzymic and non-enzymic reactions, the claim persists (Boyd, 1985) that the mechanism is concerted. The recent experimental evidence will therefore be briefly reviewed. The stepwise process for p-lactam cleavage prior to loss of the C(3) leaving group generates the enamine [36] before the conjugated imine [37]. nmr observations of the ammonolysis of cephamycins in liquid ammonia at - 50°C are consistent with the intermediate formation of the enamine [36] (Grabowski et al., 1985). The p-lactamase catalysed hydrolysis of cephalosporins shows spectral changes in the ultraviolet which are consistent with the formation of [36] prior to expulsion of the leaving group at C(3’) (Faraci and Pratt, 1984, 1985; Agathocleous et al., 1984). There is a non-linear dependence of the rate of aminolysis of cephplosporins upon hydroxide ion concentration (Page and Proctor, 1984) which is consistent with a change in rate-limiting step and hence formation of an intermediate as described in Section 11. The uncatalysed pathway k, in the rate law (8) could represent either a purely uncatalysed reaction of amine and cephalosporin, or solvent catalysis with water acting as a general base to remove a proton from the attacking amine, or as a general acid to donate a proton to the p-lactam nitrogen. The k, pathway cannot represent rate-limiting formation of the tetrahedral


251

intermediate because the observed rate constants are much smaller than the miscroscopic rate constants calculated for this step at high hydroxide ion of 1.05 for the concentration. This is substantiated by a Brernsted ,p, uncatalysed aminolysis of cephaloridine which indicates that the reaction behaves as if a unit positive charge is developed on the amine nitrogen in the transition state. The uncatalysed pathway therefore represents a ratelimiting breakdown of the tetrahedral intermediate [62] or general acid catalysed breakdown of the same intermediate by water [63]. The observed rate constant k, for the uncatalysed aminolysis must be given by k,K = k,k,/k- (Scheme 8). Since the values of K, the equilibrium constant for formation of the tetrahedral intermediate, are known, the values of k, the rate constant for the breakdown of the intermediate to products, can be calculated. These are all about lo%-' (Page and Proctor, 1984) whether or not a leaving group is expelled at C(3'). This therefore argues against expulsion of the leaving group at C(3') being concerted with fission of the carbon-nitrogen bond of the p-lactam [67]. At least this is true

~671

in the sense that there can be no significant coupling of the processes so that there is a significant lowering of the activation energy for carbon-nitrogen bond fission in the p-lactam because a group at C(3') is expelled. It could be the case that the intermediate formed by p-lactam carbonyl carbon-nitrogen bond fission is so unstable and that its lifetime is so short as to preclude its existence and therefore the breakdown of the tetrahedral intermediate is enforced to be concerted. However, the important conclusion is that expulsion of a leaving group at C(3') does not signiJcantly enhance the rate of carbon-nitrogen fission in the /l-lactam. Finally, the enamine [36] and conjugated imine [37] are in equilibrium. The imine [37] can be generated from the 0-lactamase catalysed hydrolysis of a variety of cephalosporins in aqueous solution. The addition of thiols generates the enamine [36] which can also be generated by the hydrolysis of the corresponding cephalosporin with the same thiol group at C(3'). The equilibrium between [36] and [37] is pH-dependent with the enamine favoured at low pH, (Buckwell et al., 1986). The equilibrium constant between [36] and [37] varies with the pK,-value of the thiol. The equilibrium [36][RS-][H+]/[37] shows a Brernsted PI, of 0.7, with, for example, that for butanethiol being 2 x 10-'sM2 at 30°C (Buckwell and Page, 1986).

252

MICHAEL I. PAGE

Thiols presumably add to the conjugated imine [37] by a type of Michael addition reaction. It is not inconceivable that this could be an important pathway for inactivation of enzymes by cephalosporins using a suitably placed nucleophilic group on the enzyme. 13 Reaction with alcohols and other oxygen nucleophiles

Several bacterial penicillin-binding proteins have been shown to be serine enzymes (Section 2). P-Lactamases are efficient and clinically important enzymes which play an important part in bacterial resistance to the normally lethal action of P-lactam antibiotics. A major class of p-lactamases are also serine enzymes that function by covalent catalysis with the intermediate formation of an acyl-enzyme (Knott-Hunziker et al., 1982; Cohen and Pratt, 1980; Fisher et al., 1980; Anderson and Pratt, 1983; Cartwright and Fink, 1982, Joris et al., 1984). The p-lactamase catalysed hydrolysis of a penicillin thus proceeds by formation of an acyl-enzyme which is an a-penicilloyl ester of a serine residue. The mechanism of reaction of penicillin with alcohols is therefore of obvious relevance, but in addition acyl transfer from nitrogen to oxygen nucleophiles is of current interest. The observed pseudo first-order rate constant for the degradation of benzylpenicillin in water in the presence of alcohols is given by (9), where K, kabc= k,[OH

-1

+ k,

*

4

[H+I + K ,

~

-

-

*

[ROH],

=

k,[OH-]

+ k,[RO-]

(9)

is the ionisation constant of the alcohol, k, and k, are the second-order rate constants for alkaline hydrolysis and the reaction of the alkoxide anion with penicillin, respectively. The values of k , given in Table 6 form the Brlernsted plots of Figs 10 and 14 (Davis ef al., 1987). Oxygen anions can catalyse the hydrolysis of penicillins by either acting as general bases or as nucleophiles. Several observations indicate that basic anions act as nucleophilic catalysts forming an intermediate penicilloyl ester [68] whereas weakly basic ones act as general base catalysts.

1681

The non-linear Brlernsted plot (Fig. 10) is indicative of a change in reaction mechanism. Weakly basic oxygen anions probably act as general base catalysts for the 'hydrolysis of penicillins as discussed earlier (Section 7). The


253

Bronsted p-value of 0.39 is typical for this sort of mechanism (Oakenfull and Jencks, 1971). Basic oxygen anions of pK, > 9 act as nucleophilic catalysts for hydrolysis with the intermediate formation of a penicilloyl ester [68]. The reaction of

9

10

I1

I2 13 14 pK.I ROH I

IS

16

FIG. 14 Brnnsted plot (solid line) for the second-order rate constants for the alcoholysis of benzylpenicillin against pK, for the alcohol (+). The broken line is for the second-order rate constant for the alkaline hydrolysis of a-penicilloyl esters [68]

various phenolate anions, carbohydrates and other hydroxyl-containing compounds with penicillins has been shown to proceed via the unstable ester by the penamaldate assay (Schwartz and Delduce, 1969; Schneider and DeWeck, 1968; Bundgaard, 1976d; Larsen and Bundgaard, 1978b; Yamana et al., 1977; Schwartz and Pflug, 1967; Tutt and Schwartz, 1971). The phosphate catalysed hydrolysis of benzylpenicillin is thought to occur by the nucleophilic pathway (Bundgaard and Hansen, 198I). Carbohydrates such as glucose and fructose form intermediate penicilloyl esters in their catalysed hydrolysis of penicillin and cephalosporins (Bundgaard and Larsen, 1983). It has been shown that the addition of amine nucleophiles to the p-lactam carbonyl carbon to form a tetrahedral intermediate is reversible. The rate-limiting step in the aminolysis of p-lactam antibiotics involves proton removal from the attacking amine for base catalysed reactions and p-lactam C-N bond fission for the uncatalysed reaction (Page, 1984a; Section 1 I). The rate-limiting step for the hydroxide ion catalysed hydrolysis of (j-lactams appears to be formation of the tetrahedral intermediate (p. 199).

2 54

MICHAEL I PAGE

The addition of an alkoxide ion to p-lactams generates the tetrahedral intermediate [69]. The rate-limiting step in the alcoholysis of penicillins depends on the relative rates of alkoxide ion expulsion from [69] compared with the subsequent step leading to p-lactam C-N fission. In the aminolysis of penicillins, expulsion of the attacking amine is faster than the subsequent step (p. 234). The relative rates of bond cleavage expelling alkoxide ions, compared with amines of equal basicity, are not easy to predict. Amines are expelled 105-fold faster than alkoxide ions from phthalimidium cation adducts (Gravitz and Jencks, 1974). However, the expulsion of trimethylamine from formaldehyde carbinolamine zwitterion (Hine and Kokesh, 1970) is about 104-fold slower than the expulsion of alkoxide ions, of similar basicity, from formaldehyde hemiacetal anion (Funderburk et al., 1978). In the forward direction the breakdown of the aminolysis tetrahedral intermediate [50] to products occurs with a rate constant of ca lo6 s - ’ (Gensmantel and Page, 1979a). However, this generates an unstable N-protonated amide and it is therefore expected that the equivalent reaction of [69] to generate an ester [68] will be faster because of the “push” provided by the developing conjugation from the oxygen lone pair. It would not be surprising therefore, if the rate-limiting step for the alcoholysis of p-lactams is formation of the tetrahedral intermediate. However, the large Brsnsted pnuc of 0.95 for alkoxide ions is indicative of rate-limiting breakdown of the tetrahedral intermediate. A p,,,-value of 1.0 has also been reported for the reaction of benzylpenicillin with phenolate anions (Bundgaard, 1976d; Yamana e f a/., 1977). The Brsnsted p,,, for the formation of the tetrahedral intermediate for amines adding to the (j-lactam carbonyl is 0.3 (Gensmantel and Page, 1979a). The p,,,-values for rate-limiting breakdown of tetrahedral intermediates formed from oxygen anions and esters, thioesters and acetylimidazole are 1.4, 1.4 and 1.3 respectively (Hupe and Jencks, 1977; Oakenfull and Jencks, 1971). Although the value of 1.0 is perhaps a little low compared with these values and the maximum of I .7, it is consistent with rate-limiting breakdown and is therefore yet another illustration that p-lactam C-N bond fission is not a facile process. If the reaction of benzylpenicillin with alkoxide ions proceeded with rate-limiting attack, basic alkoxide ions would be expected to show a negative deviation from the Brsnsted plot which is usually curved for such reactions (Hupe and Jencks, 1977; Jencks et al., 1982). As for the uncatalysed aminolysis of penicillins the reaction with oxygen anions could involve p-lactam C---N bond fission with or without assistance from general acid catalysis by water. The solvent isotope effect kR0,,,/kRo & for trifluoroethoxide ion reacting with benzylpenicillin is 3.9 (Davis rt a/., 1987). Although there are difficulties in the quantitative interpretation of solvent isotope effects (Gold and Grist, 1972) the values of


255

0.5-4.7 for the attack of methoxide ion on phenyl acetates and methyl

phenyl carbonates in MeOD are typical for rate-limiting formation of tetrahedral intermediates (Mitton et al., 1969). It seems likely therefore that the mechanism for the formation of penicilloyl esters involves general acid catalysis by the solvent as shown in [70]. No term for buffer catalysis is observed in the rate law for alcoholysis of penicillin (Davis et al., 1987); this would be expected if buffers could act as general acid catalysts.

[701

It is interesting to note that the hydroxide ion catalysed hydrolysis of benzylpenicillin involves rate-limiting attack whereas alkoxide ions react with benzylpenicillin with rate-limiting breakdown of the tetrahedral intermediate. This could result from the intermediate formed by hydroxide ion attack either breaking down to reactants or products slower or faster, respectively, than [69]. The product of the alkoxide-ion catalysed hydrolysis of benzylpenicillin is benzylpenicilloic acid (see, however, Section 14). The hydrolysis of the intermediate a-penicilloyl esters [68] can be studied independently (Davis et al., 1986). The Brernsted &-value for the hydroxide-ion catalysed hydrolysis of the esters is -0.3, electron withdrawing substituents increasing the rate of hydrolysis (Fig. 14). Consequently, the oxygen anion catalysed hydrolysis of benzylpenicillin does not give a detectable ester intermediate for alcohols of pKa < 12. Conversely, for alcohols of pKa > 14 the hydrolysis of the ester intermediate is slower than its rate of formation. It is interesting to note that the rates of hydrolysis of the ester intermediates show no evidence of steric inhibition by the penicilloyl residue and are similar to those of simple alkyl esters (Davis et al., 1986; Gensmantel et al., 1981). This suggests that, if slow hydrolysis of the penicilloyl enzyme formed during antibiosis is responsible for enzyme inhibition, it is not the result of intrinsic stabilisation but the result of interaction with the protein preventing access to the ester function. Penicilloyl esters of basic alcohols undergo reactions in addition to hydrolysis (Davis and Page, 1985) and these are discussed below.

THIAZOLIDINE RING OPENING

The first chemical step in the antibacterial activity of penicillins is thought to be the ring opening of the p-lactam by a serine hydroxyl of a transpeptidase

MICHAEL I. PAGE

256

enzyme to form an intermediate which is an ester of penicilloic acid [68; R = Enz] (Waxman and Strominger, 1980). It is not known if it is the stability of this intermediate which causes enzyme inhibition and why the ester is not hydrolysed to regenerate the enzyme and penicilloic acid. It is conceivable that a subsequent reaction of the ester produces an electrophilic entity which is ultimately responsible for enzyme inhibition by irreversibly reacting with a nucleophilic group on the enzyme. Methyl 5R, 6R-benzylpenicilloate [68; R = Me] in water undergoes reactions other than simple ester hydrolysis and produces intermediates which may be of relevance to enzyme inhibition. The alkaline hydrolysis of methyl 5R, 6R-benzylpenicilloate shows an optical density increase followed by a decrease at 280 nm with the absorbance maximum increasing with concentration of hydroxide ion. The observed pseudo first-order rate coefficients kobs for both of these two phases show non-linear dependences upon hydroxide ion. For the first phase kobs is pH-independent up to 0.07M sodium hydroxide but at higher concentrations it becomes first-order in hydroxide ion. By contrast, kobs for the second phase, corresponding to a decrease in optical density, changes from being hydroxide-ion dependent to independent with increasing pH. These kinetic observations are compatible with Scheme 10 where SH = ester [68; R = Me]. The only observable products are penicilloic acid [30] and methanol (Davis and Page, 1985). The ester [68; R = Me] undergoes base-catalysed reversible ionisation in a pHdependent equilibrium and at a rate which is faster than the hydroxide-ion catalysed hydrolysis of the ester. If ionisation corresponds to a simple deprotonation of [68] then it would have an apparent pK, of 12.9 and k,, the second-order rate constant, is 0.55 M-'s-', as expected for hydrolysis of a methyl ester. k,

OH-+SH

I

+ I SA ,

hz

Products Scheme 10

The simplest interpretation of these observations is that methyl 5R, 6R-benzylpenicilloate [68] undergoes reversible elimination across C(6bC(5) and ring opening of the thiazolidine to give the enamine tautomer of methyl penamaldate [71] (Scheme 1 I ) . The amount of enamine formed increases with increasing hydroxide concentration which explains the increasing absorbance at 280 nm, the wavelength of the absorption maxima of penamaldates. The thiol anion can be trapped with Ellman's reagent. Hydrolysis

257


H H

_...

RCOHNn

MeO,

sTCH3 "CH, H

-

RCONH

L

.-co; H

HH H

-0,C

N H

H H RCON

Me0,C

"COT

HY RCON

-0,c

co;

co;

co;

H H RCON

H

-0,c

N

H

Scheme I I

of the ester occurs at low concentrations of hydroxide ion by the normal base-catalysed mechanism but becomes pH-independent at high concentrations of hydroxide because the substrate exists predominantly as the

258

MICHAEL I. PAGE

enamine tautomer [7 I], which is presumably less reactive towards ester hydrolysis. Hydrolysis of the ester must occur through [68], the concentration of which decreases with increasing pH (Davis and Page, 1985). R'CONH

S.-

[711

[721

Similar observations have been made with 6-a-halopenicillanic acids but in this case the thiol anion in the analogous enamine [71] displaces the halide at C(6) via the imine tautomer [72] (Gensmantel et al., 1981). If the enamine [7 I] is produced in the transpeptidase-catalysed reaction with penicillins, then inactivation of the enzyme could occur by a nucleophilic group on the enzyme attacking C(5) in a Michael-type addition reaction. 14 Epimerisation of penicillin derivatives

The initial product of alkaline hydrolysis of benzylpenicillin is 5R, 6Rbenzylpenicilloic acid. However, epimerisation then occurs at C(5) to give a mixture of the 5R, 6R- and 5S, 6R- penicilloic acids [73] (Davis and Page, 1985; Carroll et al., 1977; Busson et al., 1976b; Kessler et al., 1983; Bird et al., 1983). This change in configuration at C(5) is accompanied by a decrease in pK, of the protonated thiazolidine from 5.3 to 4.8, a change in the nmr chemical shifts of the protons at C(5) and C(6) and those on the a and p methyl groups at C(2), a decrease in the coupling constants between the C(5)-C(6) hydrogens and a change in specific rotation. Epimerisation does not occur at C(6). The equilibrium constant for the ratio of the 5S, 6R-epimer to that of the 5R, 6R-benzylpenicilloate is 4. The observed polarimetric pseudo first-order rate constants for this process are pH- and buffer-independent from pH 6 to pH 12.5 but become first order in hydroxide ion at higher pH. Epimerisation in D,O occurs without D-incorporation at C(5) or C(6) over most of the pH range (Davis and Page, 1985).


259

The pH-independent epimerisation of penicilloic acid at C(5) probably occurs by unimolecular ring opening of the thiazolidine to form the iminium ion tautomer of penamaldic acid [74]. There are two possible explanations for the base-catalysed reaction. Above pH 12 deprotonation of the iminium ion could become faster than ring closure, and the rate of epimerisation would increase because the steady state concentration of the ring-opened thiazolidine is increased. Alternatively, epimerisation could occur by the penamaldate mechanism followed by penicilloyl esters (Davis and Page, 1985). At hydroxide ion concentrations above 1 M elimination across C(stC(5) and thiazolidine ring-opening occurs to give the enamine, similar to [71]. This is suggested by the observation of the appearance of the chromophore at 280 nm and deuterium-exchange at C(6) (Davis and Page, 1986). Around neutral pH a-penicilloyl esters [68] epimerise at C(5) faster than the ester function is hydrolysed (Davis and Page, 1985). If this occurs with penicilloyl enzymes from penicillins and the serine hydroxyl at the active site of the enzyme, the ring-opening of the thiazolidine generates electrophilic sites capable of irreversibly inactivating the enzyme. Penicillins with common amido side chains at C(6) are hydrolysed initially to 3S, 5R, 6R-penicilloic acid without D-incorporation or epimerisation at C(3), C(5) or C(6). However, penicillins without an ionisable NH on the C(6) side chain can undergo epimerisation from 6 4 - to 6-a-substituted penicillins faster than ring-opening of the p-lactam (Wolfe and Lee, 1968; Johnson and Mania, 1969). The epimerisation of hetacillins to epihetacillin and that of mecillinam is hydroxide-ion catalysed but shows no general base catalysis (Tsuji et al., 1977; Baltzer et al., 1979). There have been several unusual observations related to proton abstraction at C(6) in penicillins which indicate that the carbanion formed is exceptional. Treatment of the penicillin Schiff base derivative [75] with phenyllithium in tetrahydrofuran appears to lead to abstraction of the C(6) proton but reprotonation regenerates [75] despite the 6-a-epimer being thermodynamically more stable (Firestone et al., 1972, 1974). Even more strange is the observation that reprotonation with D,O/CD,CO,D does not give the 6-a-deuterio Schiff base [75]. However, epimerisation by triethylamine in acetonitrile containing D,O is accompanied by deuteriation of the ArCH=N

0 ’

H3

CO,CH,Ph

P‘51

X

D,

0.’

[761

P-prot o n a t i o n ci-protonation

6-p-X

6-a-X

FIG. 15 Energy-profiles for abstraction of the proton on C(6) in 6-X-substituted penicillins and reprotonation of the intermediate carbanion


261

presumed carbanion intermediate [76] and occurs preferentially from the least hindered a-face to give the less stable P-epimer. Deuterium exchange at C(6) of the 6 4 epimer occurs faster than epimerisation at C(6) (Firestone and Christensen, 1977). The calculated isotope effect k,/k, for protonation of the carbanion is 12.7 which is ascribed to proton tunnelling. By contrast, 6-a-chloropenicillanic acid undergoes D-exchange at C(6) in NaOD/D,O without epimerisation and at a rate faster than P-lactam ring-opening. 6-P-Epimers appear to undergo epimerisation and deuteriation at C(6) at the same rate in water (Clayton et al., 1969; Gensmantel et al., 1981). If it is assumed that epimerisation and D-exchange at C(6) occur via the same carbanion [76], two schemes can be envisaged (Fig. 15). In ( a ) P-protonation of the carbanion is faster than a-protonation whereas the reverse is true in (6). P-Protonation would have a lower activation energy if either the transition state is “late” and the factors stabilising the a-epimer in the product are reflected in the transition state [presumably less unfavourable steric interactions between the C(6) substituent and the C(2) methyl] or an “early” transition state but a non-planar carbanion [76] with a preponderance of electron density on the P-face (Gensmantel et a / . , 1981). For a planar carbanion [76] an “early” transition state for protonation would favour the sterically favourable exo a-direction. It is conceivable that there is a change from ( a ) to (6) (Fig. 15) on going from aqueous to non-polar solvents. As described in the previous sections protonation of the enamine [71] appears to occur stereospecifically since the stereochemistry at C(6) is preserved when [7I] ring closes to regenerate [68]. There have been several reports of the rates and mechanism of epimerisation of asymmetric centres in side chains of p-lactam antibiotics. These have usually involved an asymmetric centre adjacent to the carbonyl carbon of the penicillin C(6) or cephalosporin C(7) amide side chain (Bird et al., 1984; Hashimoto and Tanaka, 1985). References

Abraham, E. P., Chain, E., Baker, W. and Robinson, R. (1943). Pen Report No. 103 Abrahamsson, S., Hodgkin, D. C. and Maslen, E. N. (1963). Biochem. J . 86, 514 Adriaens, P., Meesschaert, B., Frere, J. M., Vanderhaeghe, H., Degelaen, J., Ghuysen, J . M. and Eyssen, H. (1978). J . Biol. Chem. 253, 3660. Agathocleous, D., Buckwell, S., Proctor, P. and Page, M. I. (1985). I n “Recent Advances in the Chemistry of p-Lactam Antibiotics” (eds A. G . Brown and S . M. Roberts) p. 18. Royal Society of Chemistry, London Allinger, M. L., Tribble, M. T. and Miller, M . A. (1972). Tetrahedron 28, 1173 Anderson, B. M., Cordes, E. H. and Jencks, W. P. (1961). J . Biol. Chem. 236, 455 Anderson, E. G. and Pratt, R. F. (1983). J . Biol. Chem. 258. 13120

262

MICHAEL I. PAGE

Anoardi, L. and Tonellato, V. (1977). J . Chem. SOC.Chem. Commun. 401 Applegate, H. E., Kolfini, J. E., Puar, M. S., Slusarchyk, W. A., Toeplitz, B. and Gougoutas, J. Z . (1974). J . Org. Chem. 39, 2794 Aroney, M. J., Le Fere, R. J. W., Radom, L. and Ritchie, G . L. D. (1968). J . Chem. SOC.B 507 Asso, M., Panossian, R. and Guiliano, M. (1984). Spectroscopy Lett. 17, 271 Baltzer, B., Lund, F. and Rastrup-Andersen, N. (1979). .I. Pharm. Sci. 68, 1207 Batchelor, F. R., Dewdney, J. M. and Gazzard, D. (1965). Nature 206, 362 Bender, M. L. and Thomas, R. J. (1961). J. Am. Chem. SOC.83, 4183 Berge, S. M., Henderson, N. L. and Frank, M. J. (1983). J . Pharm. Sci. 72, 59 Berger, S. A. (1984). New England J . Med. 312, 1395. Bird, A. E., Cutmore, E. A., Jennings, K. R. and Marshall, A. C. (1983). J . Pharm. Pharmacol. 35, 138 Bird, A. E., Charsley, C.-H., Jennings. K. R. and Marshall, A. C. (1984). Analyst 109, 1209 Blackburn, G . M. and Dodds, H. L. H. (1974). J . Chem. SOC.Perkin Trans 2 377 Blackburn, G . M. and Plackett, J. D. (1972). J. Chem. SOC.,Perkin Trans 2 1366 Blackburn, G . M. and Plackett, J. D. (1973). J. Chem. SOC.,Perkin Trans 2 981 Blaha, J. M., Knevel, A. M., Kessler, D. P., Mincy, J. W. and Hern, S. L. (1976). J . Pharm. Sci. 65, 1165 Bolton, P. D. and Jackson, G . L. (1969). Austral. J. Chem. 22, 527 Bose, A. K. and Srinivasan, P. R. (1979). Org. Mag. Res. 12, 34 Bottini, A. T. and Nash, C. P. (1962). J . Am. Chem. SOC.,84, 734 Bouchoux, G. and Houriet, R. (1984). Tetrahedron Lett. 5755 Boyd, D. B. (1972). J . Am. Chem. SOC.94, 6513 Boyd, D. B. (1982). In “Chemistry and Biology of 0-Lactam Antibiotics”, (eds R. B. Morin and M. Gorrnan) Vol. 1, p. 437 Academic Press, New York Boyd, D. B. (1983). J . Med. Chem. 26, 1010 Boyd, D. B. (1985). J . Org. Chem. 50, 886 Boyd, D. B. and Lunn, W. H. W. (1979). J. Med. Chem 22, 778 Boyd. D. B., Hermann, R. B., Presti, D. E. and Marsh, M. M. (1975). J . Med. Chem. 18,408 Boyd, D. B., Yeh, C.-Y. and Richardson, F. S. (1976). J . Am. Chem. SOC.98, 6100 Boyd, D. B., Herron, D. K., Lunn, W. H. W. and Spitzer, W. A. (1980). J . Am. Chem. SOC.102, 1812. Brown, H. C. and Ichikawa, K. (1957). Tetrahedron 1, 221 Broxton, T. J. and Duddy, N. W. (1979). Austral. J . Chem. 32, 1717 Bruylants, A. and Kezdy, F. (1960). Rec. Chem. Progr. 21, 213 Buckwell, S. and Page, M. I . (1985). Unpublished observations Buckwell, S., Page, M. I. and Longridge, J. L. (1986). J . Chem. SOC.Chem. Commun. 1039 Bucourt, R., Heymes, R., Lutz, A., Penasse, L. and Perronet, J. (1978). Tetrahedron 34,2233 Bundgaard, H. (1971a). J . Pharm. Sci. 60, 1273 Bundgaard. H. (1971b). Tetrahedron Lett. 4613 Bundgaard, H. (1972). Dansk. Tidsskr. Farm. 46, 29, 85 Bundgaard, H. (1973). Acta Pharm. Suecica 10, 309 Bundgaard. H. (1975). Arch. Pharm. Chemi. Sci. Ed. 3, 94 Bundgaard, H. (1976a). J . Pharm. Pharmacol. 28, 725 Bundgaard, H. (1976b). Arch. Pharm. Chemi. Sci. Ed. 4, 25 Bundgaard, H. (1976~).Arta. Pharm. Suecica 13, 299


263

Bundgaard, H. (1976d). Arch. Pharm. Chemi. Sci. Ed. 4,91 Bundgaard, H. (1977a). Acta Pharm. Suecica 14,47, 67 Bundgaard, H. (1977b). Arch. Pharm. Chemi. Sci. Ed. 5, 141, 149 Bundgaard, H. (1980). Arch. Pharm. Chemi. Sci. Ed. 8, 161 Bundgaard, H. and Angelo, H. R. (1974). Tetrahedron Lett. 3001 Bundgaard, H. and Buur, A. (1983). Arch. Pharm. Chem. Sci. Ed. 11, 15 Bundgaard, H. and Hansen, J. (1981). Int. J. Pharm. 9,273 Bundgaard, H. and Ilver, K. (1970). Dansk. Tidsskr. Farm. 44, 365 Bundgaard, H. and Larsen, C. (1983). Int. J. Pharm. 16, 319 Bunton, C. A. (1977). Pure Appl. Chem. 49,969 Bunton, C. A. (1984). In “The Chemistry of Enzyme Action” (ed. M. I. Page) p. 461. Elsevier, Amsterdam Bunton, C. A. and Savelli, G. (1986), Adv. Phys. Org. Chem. 22, 213 Bunton, C. A., Robinson, L. and Stam, M. (1970). J. Am. Chem. SOC.92,7393 Buss, V. Gleiter, R. and Schleyer, P. von R. (1971). J. Am. Chem. SOC.93,3927 Busson, R. and Vanderhaeghe, H. (1978). J. Org. Chem. 43, 4438 Busson, R., Claes, P. J. and Vanderhaeghe, H. (1976a). J. Org. Chem. 41,2556 Busson, R., Vanderhaeghe, H. and Toppet, S. (1976b). J. Org. Chem. 41, 3054 Busson, R., Roets, E. and Vanderhaeghe, H. (1978). J. Org. Chem. 43,4434 Butler, A. R., Robinson, I. R. and Wright, D. E. (1982). J. Chem. SOC.Perkin Trans2 827 Carlsen, N. R., Radom, L. and Riggs, N. V. (1979). . I Am. . Chem. SOC.101, 2233 Carroll, R. D., Jung, S. and Sklavounos, C. G. (1977). J. Heterocycl. Chem. 14,503 Cartwright, S. J. and Fink, A. L. (1982). FEBS Lett. 137, 186 Chaimovich, H., Correia, V. R., Araujo, P. S., Aleixo, M. V. and Cuccovia, I. M. (1985). J . Chem. SOC.Perkin Trans 2 925 Chakrawarti, P. B., Tiwari, A. and Sharma, H. N. (1982). Indian J. Chem. 21A,200 Chakrawarti, P. B., Tiwari, C. P., Tiwari, A. and Sharma, H. N. (1984). J. Indian Chem. SOC.61, 705 Chambers, R. and Doedens, R. J. (1980). Acta Crystallogr. Sect. B. 36, 1507 Clayton, J. P., Nayler, J. H. C., Southgate, R. and Stove, E. R. (1969). J. Chem. SOC. Chem. Commun. 129 Coene, B., Schanck, A., Dereppe, J.-M. and Van Meerssche, M. (1984). J. Med. Chem. 27, 694 Cohen, S. A. and Pratt, R. F. (1980). Biochemistry 19, 3996 Collings, A. J., Jackson, P. F. and Morgan, K. J. (1970). J. Chem. SOC.B 581 Cordes, E. H. (1978). Pure Appf. Chem. 49,969 Cressman, W. A., Sugita, E. T., Coluisio, J. T. and Niebergall, P. J. (1969). J. Pharm. Sci. 58, 1471 Crowfoot, D., Bunn, C. W., Rogers-Low, B. W. and Turner-Jones, A. (1949). In “The Chemistry of Penicillin” (eds H. T. Clarke, J. R. Johnson and R. Robinson) p. 310. Princeton University Press, Princeton, New Jersey Davis, A. M. and Page, M. I. (1985). J . Chem. SOC.Chem. Commun. 1702 Davis, A. M . and Page, M. I. (1986). Unpublished observations Davis, A. M., Proctor. P. and Page, M. I. (1987). J . Chem. SOC.Perkin Trans 2 In press Degelaen, J., Loukas, S. L., Feeney, J., Roberts, G. C. K. and Burgen, A. S. V. (1979a). J. Chem. SOC.Perkin Trans 2 86 Degelaen, J., Feeney, J., Roberts, G. C. K., Burgen, A. S. V., Frtre, J. M. and Ghuysen, J. M . (1979b). FEBS Lett. 98,53

264

MICHAEL I. PAGE

Demarco, P. V. and Nagarajan, R. (1973). In “Cephalosporins and Penicillins” (ed. E. H. Flynn) Ch. 8. Academic Press, New York Dennen, D. W. and Davis, W. W. (1962). Antimicrob. Ag. Chemother. 1, 531 Dereppe, J. M., Schanck, A., Coene, B., Moreau, C. and Van Meerssche, M. (1978). Org. Magn. Resonance 11, 638 Deslongchamps, P. (1975). Tetrahedron 31, 2463 DeWeck, A. L. (1962). Int. Arch. Allergy Appl. Immunol. 21, 20 DeWeck, A. L. and Bulm, G. (1965). Int. Arch. Allergy Appl. Immunol. 27, 221 Dewhirst, K. C. and Cram, D. J. (1958). J. Am. Chem. SOC.80, 31 15 Dhami, K. S. and Stothers, J. B. (1964). Tetrahedron Lett. 631 Dinner, A. (1977). J. Med. Chem. 20, 963 Doering, W. von E., Birladeanu, L., Andrews, D. W. and Pagnotta, M. (1984). J. Am. Chem. SOC.107, 428 Doyle, F. P. and Nayler, J. H. C. (1964). In “Advances in Drug Research” (eds N. J. Harper and A. B. Simmonds) Vol. I , p. 1. Academic Press, New York Dusart, J., Marquet, A,, Ghuysen, J. M., Frere, J. M., Moreno, R., Leyh-Bouille, M., Johnson, K., Lucchi, C., Perkins, H. R. and Nieto, M. (1973). Antimicrob. Agents Chemo. Ther. 3, 181 Earl, H. A,, Marshall, D. R. and Stirling, C. J. M. (1983). J. Chem. SOC.Chem. Commun. 779 Eichelberger, H. R., Mayeste, R. J. and Good, M. L. (1974). 16th Proc. Int. Con$ Coord. Chem. 1, 21 Eichorn, G. (1973). Inorg. Biochem. 1, 79 Eigen, M. (1964). Angew. Chem. [Int. Edn] 3, 1 Faraci, W. S. and Pratt, R. F. (1984). J. Am. Chem. SOC.106, 1489 Faraci, W. S. and Pratt, R. F. (1985). Biochemistry 24, 903 Fazakerley, G . V. and Jackson, G . E. (1975a). J. Chem. SOC.Perkin Trans 2 567 Fazakerley, G. V. and Jackson, G. E. (1975b). J . Inorg. Nucl. Chem. 37, 2371 Fazakerlev. G . V. and Jackson, G . E. (1977). J. Pharm. Sci. 66. 533 Fazakerle;, G . V. Jackson, G. E. and Linder; P. W. (1976). J. Inorg. Nucl. Chem. 38, 1397 Fendler, J. H. and Fendler, E. J. (1975). “Catalysis in Micellar and Macromolecular Systems”. Academic Press, New York Fersht, A. R. and Requena, Y. (1971). J. Am. Chem. SOC.93, 3499, 3502 Finholt, P., Jurgensen, G. and Kristiansen, H. (1965). J. Pharm. Sci. 54,387 Firestone, R. A. and Christensen, 9. G. (1977). J. Chem. SOC.Chem. Comrnun. 288 Firestone, R. A,, Schelechow, N., Johnston, D. B. R. and Christensen, B. G. ( 1 972). Tetrahedron Lett. 375 Firestone, R. A,, Maciejewicz, N. S., Ratcliffe, R. W. and Christensen, B. G . (1974). J . Org. Chem. 39, 437 Firestone, R. A., Fahey, J. L., Maciejewicz, N. S., Patel, G. S. and Christensen, B. G. (1977). J . Med. Chem. 20, 551 Fisher, J., Belasco, J. G., Khosla, S. and Knowles, J. R . (1980). Biochemistry 19,2985 Fleming, A. (1929). Br. J. Exp. Pathol. 10, 226 Flynn, E. H. (1972). “Cephalosporins and Penicillins: Chemistry and Biology” Academic Press, New York Frere, J. M. and Joris, B. (1985). CRC‘ Crit. Revs. Microbiol. 11, 289 Frere, J. M., Leyh-Bouille, M., Ghuysen, J. M. and Perkins, H. R. (1974). Eur. J. Biochem. 50. 203 Frere, J. M.. Ghuysen, J. M. and Iwatsubo, M. (1975a). Eur. J. Biochem. 57, 343

MECHANISMS OF REACTIONS OF B-LACTAM ANTIBIOTICS

265

Frere, J. M., Ghuysen, J. M., Degelaen, J., Loffet, A. and Perkins, H. R. (1975b). Nature (London) 258, 168 Frere, J. M., Ghuysen, J. M., Vanderhaeghe, H., Adriaens, P., Degelaen, J. and De Graeve, J. (1976a). Nature (London) 260. 451 Frere, J. M., Duez, C., Ghuysen, J. M. and Vandekerkhove, J. (1976b). FEBS Lett. 70, 727 Frere, J. M., Ghuysen, J. M. and De Graeve. J. (1978). FEBS Lett. 88. 147 Frere, J. M., Kelly, J. A., Klein, D., Ghuysen, J. M. Claes, P. J. and Vanderhaeghe, H. (1982). Biochem. J . 203, 223 Frere, J. M., Klein, D., Kelly, J. A. and Ghuysen, J. M. (1984). FEMS Microbiol. Lett. 21 213 Fujita, T. and Koshiro, A. (1984). Chem. Pharm. Bull. 32, 3651 Funderburk, L., Aldwin, L. and Jencks, W. P. (1978). J . Am. Chem. SOC.100. 5444 Gani, V. and Viout, P. (1978). Tetrahedron 34, 1337 Gassman, P. G., Cryberg, R. L. and Shudo, L. (1972). J . Am. Chem. SOC.,94, 7600 Gensmantel, N. P. and Page, M. I . (1979a). J . Chem. SOC.Perkin Trans 2 137 Gensmantel, N. P. and Page, M. I . (1979b). Unpublished observations Gensmantel, N. P. and Page, M. I. (1982a). J . Chem. Soc. Perkin Trans 2 147 Gensmantel, N. P. and Page, M. I. (1982b). J . Chem. SOC.Perkin Trans 2 155 Gensmantel, N. P., Cowling, E. W. and Page, M. I. (1978). J . Chem. Soc. Perkin Trans 2 335 Gensmantel. N. P., Proctor, P. and Page, M. I . (1980). J . Chem. SOC.Perkin Trans 2 1725 Gensmantel, N. P., McLellan, D., Morris, J. J., Page, M. I., Proctor, P. and Randahawa, G. (1981). In “Recent Advances in the Chemistry of p-Lactam Antibiotics” (ed. G. Gregory) p. 227. Royal Society of Chemistry, London Georgopapadakou. N. H., Smith, S. A. and Cimarusti, C. M. (1981). Eur. J . Biochem. 115, 53 Ghuysen, J.-M. (1977). Cell Surfaces Rev. 4. 463 Ghuysen, J.-M., Frere, J.-M., Leyh-Bouille, M., Nguyen-Disteche, M., Coyette, J.. Dusart, J., Joris, B., Duez, C., Dideberg, O., Charlier, P., Dive, G. and LamotteBrasseur, J. (1985). Scand. J . Infect. Dis. Suppl. 42, 17 Giffney, C. J. and O’Connor, C. J. (1975). J . Chem. SOC.Perkin Trans 2 1357 Gilpin, M. L., Harbridge, J. B., Howarth, T. F. and King, T. J. (1981). J . Chem. SOC. Chem. Commun. 929 Glidewell, C. and Mollison, G. S. M. (1981). J . Mol. Struct. 72. 203 Gold, V. and Grist, S. (1972). J . Chem. SOC.Perkin Trans 2 89 Grabowski, E. J. J., Douglas, A. W. and Smith, G. B. (1985). J . Am. Chem. SOC.107, 267 Gravitz, N. and Jencks, W. P. (1974). J . Am. Chem. Soc. %, 499 Green, G. G. F. H., Page, J. E. and Straniforth, S. E. (1965). J . Chem. SOC.1595 Hamilton-Miller, J. M . T., Richards, E. and Abraham, E. P. (1970a). Biochem. J . 96, 739 Hamilton-Miller, J. M . T., Richards, E. and Abraham. E. P. (1970b). Biochem. J . 116, 385 Hammond, G . S. (1956). In “Steric Effects in Organic Chemistry” (Ed. M. S. Newman), Ch. 9. John Wiley and Sons, New York Hashimoto, N. and Tanaka, H. (1985). J . Pharm. Sci. 74, 68 Hehre, W. J., Radom, L. and Pople, J. A. (1972). J . Am. Chem. SOC.94, 1496 Higgins, C. E., Hamill, R. L., Sands, T. H., Hoehn, M . M., Davis, N. E., Nagarajan, R. and Boeck, L. D. (1974). J . Antibiot. 27, 298

266

MICHAEL I. PAGE

Hine, J. and Kokesh, F. C. (1970). J. Am. Chem. SOC.92, 4383 Ho, P. P. K., Towner, R. D., Indelicato, J. M., Wilham, W. L., Spitzer, W. A. and Koppel, G. A. (1973). J . Antibiot. 36, 313 Hong, J. T. and Kostenbauder, H. B. (1975). J . Pharm. Sci. 64, 1378 Hupe, D. J. and Jencks, W. P. (1977). J. Am. Chem. SOC.99, 451 Indelicato, J. M. and Wilham, W. L. (1974). J. Med. Chem. 17, 528 Indelicato, J. M., Norvilas, T. T., Pfeiffer, R. R., Wheeler, W. J. and Wilham, W. L. (1974). J . Med. Chem. 17, 523 Irving, H. and Williams, R. J. P. (1953). J. Chem. SOC.3192 Ishidate, H., Sakaguchi, T., Taguchi, K. and Kanao, S. (1960). Anal. Chim. Acta 22, 452 James, M. N. G., Hall, D. and Hodgkin, D. C. (1968). Nature (London) 220, 168 Jeffrey, G. A., Ruble, J. R., McMullen, R. K., DeFrees, D. J., Binkley, J. S. and Pople, J. A. (1980). Acta Cryst. B36,2242 Jencks, W. P. (1975). Adv. Enzymol, 43, 219 Jencks, W. P. (1976). Accounts Chem. Res. 9, 425 Jencks, W. P. and Page, M. I. (1 972). Proc. Eighth FEBS Meeting, Amsterdam, 29,45 Jencks, W. P., Brant, S. R., Gandler, J. R., Fendrich, G. and Nakamura, C. (1982). J. Am. Chem. SOC.104, 7045 Johnson, D. A. and Mania, D. (1969). Tetrahedron Lett. 1779 Joris, B., Dusart, J., Frere, J.-M., van Beeumen, J., Emanuel, E. L., Petursson, S., Gagnon, J. and Waley, S. G. (1984). Biochem. J . 233, 271 Kelly, J. A., Frere, J.-M., Klein, D. and Ghuysen, J. M. (1981). Biochem. J . 199, 129 Kelly, J. A., Knox, J. R., Moews, P. C., Hite, G. J., Bartolone, J. B., Haiching, Z., Joris, B., Frere, J.-M. and Ghuysen, J.-M. (1985). J. Biol. Chem. 260, 6449 Kessler, H. (1970). Angew. Chem. / h i . Edn] 9, 219 Kessler, D. P., Cushman, M., Ghebre-Sellassie, I., Knevel, A. M. and Hem, S. L. (1983). J . Chem. SOC.Perkin Trans 2 1699 Kirby, A. J. (1980). Adv. Phys. Org. Chem. 17, 183 Knott-Hunziker, V., Petursson, S., Waley, S. G., Jaurin, B. and Grundstrom, T. (1982). Biochem. J . 207, 315 Komiyama, M. and Bender, M. L. (1984). In “The Chemistry of Enzyme Action” (ed. M. I. Page) p. 505. Elsevier, Amsterdam Krabbenhoft, H. A., Wiseman, J. R. and Quinn, C. B. (1974). J. Am. Chem. SOC.96, 258 Kresge, A. J., Fitzgerald, P. H. and Chiang, Y. (1974). J. Am. Chem. SOC.96, 4698 Larsen, C. and Bundgaard, H. (1977). Arch. Pharm. Chemi. Sci. Ed. 5, 201 Larsen, C. and Bundgaard, H. (1978a). Arch. Pharm. Chemi. Sci. Ed. 6 33 Larsen, C. and Bundgaard, H. (1978b). J. Chromat. 147, 143 Laurent, G., Durant, F., Frere, J. M., Klein, D. and Ghuysen, J. M. (1984). Biochem. J . 218, 933 Levine, B. B. (1961). Arch. Biochem. Biophys. 93, 50 Levine, B. B. and Ovary, Z. (1961). J . Exp. Med. 114, 875 Levy, G. C. and Nelson, G. L. (1972a). “Carbon-13 Nuclear Magnetic Resonance for Organic Chemists”. Wiley, Interscience, New York Levy, G. C. and Nelson, G. L. (1972b). J . Am. Chem. SOC.94, 4897 Lichter, R. L. and Dorman, D. E. (1976). J. Org. Chern. 41, 582 Liler, M. (1969). J. Chem. SOC.B. 385 Liler, M. (1975). Adv. Phys. Org. Chem. 11, 328 Longridge, J. L. and Timms, D. (1971). J . Chem. SOC.B. 852


267

Luche, J. L., Kagan, H. B., Parthasarathy, R., Tsoucaris, G., De Rango, C. and Zeliver, C. (1968). Tetrahedron 24, 1275 Lumbreras, J. M., Fernandez, G. M. and Ordonez, D. (1982). An. Real Acad. Farm.

48, 5 1 Malihowski, E. R., Manhas, M. S., Goldberg, M. and Fanelli, V. (1974). J. Mol. Struct. 23,32 1 Manhas. M. S., Jeng, S. and Bose, A. K. (1968). Tetrahedron 24, 1237 Martin, A. F., Morris, J. J. and Page, M. I. (1976). J. Chem. SOC.Chem. Cornmun., 495 Martin, A. F., Morris, J. J. and Page, M. I. (1979). J. Chem. SOC.Chem. Commun., 298 Martinek, K., Yatsimirski, A. K., Osipov, A. P. and Berezin, I. V. (1973). Tetrahedron 29, 963 Martinek, K., Yatsimirski, A. K., Osipov, A. P. and Berezin, I. V. (1975). Tetrahedron 31, 709 Martinek, K., Yatsimirski, A. K., Levashov, A. V. and Berezin, I. V. (1977). In “Micellisation, Solubilisation and Microemulsions” (ed. K. L. Mittal,) Vol. 2. Plenum Press, New York McLelland, R. A. and Reynolds, W. F. (1974). J. Chem. SOC.Chem. Commun. 824 McLelland, R. A. and Reynolds, W. F. (1976). Can. J. Chem. 54, 718 Menger, F. M. and Portnoy, C. E. (1967). J . Am. Chem. SOC.89,4698 Meyer, W. P. and Martin, J. C. (1976). J. Am. Chem. SOC.98, 1231 Miller, I. M., Stapley, E. 0. and Chaiet, L. (1962). Bart. Proc. A49, 32 Minhas, H. S. and Page, M. I. (1982). Unpublished observations Mitton, C. G., Gresser, M. and Schowen, R. L. (1969). J. Am. Chem. SOC.91,2045 Modro, T. A,, Beaufays, F. and Yates, K. (1977). Can. J. Chem. 55,3050 Moll, F. (1968). Arch. Pharm. 301,272 Molyneux, P., Rhodes, C. T. and Swarbick, J. (1965). Trans. Faraday SOC.,61, 1043 Mondelli, R. and Ventura, P. (1977). J. Chem. SOC.Perkin Trans 2 1749 Morin, R. B., Jackson, B. G., Mueller, R. A,, Lavagnino, E. R., Scanlon W. B. and Andrews, S. L. (1969). J. Am. Chem. SOC.91, 1401 Morris, R. B. and Jackson, B. G. (1970). Fortsch. Chem. Org. Naturst. 28, 343 Morris, J. J. and Page, M. I. (1978). Unpublished observations Morris, J. J. and Page, M. I. (1980a). J. Chem. SOC.Perkin Trans 2 212 Morris, J. J. and Page, M. I. (1980b). J. Chem. SOC.Perkin Trans 2 220 Morris, J. J. and Page, M. I. (1980~).J. Chem. SOC.Perkin Trans 2 679, 685 Murakami, K., Takasuka, M., Motokawa, K. and Yoshida, T. (1981). J. Med. Chem. 24,88 Murakami, K. and Yoshida, T. (1982). Antirnicrob. Agents Chemother. 21, 254 Nagarajan, R., Boeck, L. D., Gorman, M., Hamill, R. L., Higgens, C. E., Hoehn. M. M., Stark, W. M. and Whitney, J. G. (1971). J. Am. Chem. SOC.93, 2308 Narisada, M., Onoue, H. and Nagata, W. (1977). Heterocycles 7, 389 Narisada, M., Yoshida, T., Ohtani, M., Ezumi, K. and Takasuka. M. (1983). J. Med. Chem. 26, 1577 Nelson, H. D. and DeLigny, D. L. (1968). Red. Trav. Chim. Pays-Bas 87,623 Newton, G. G. F. and Abraham, E. P. (1955). Nature (London) 175, 548 Nikaido, H. (1981). In “P-Lactam Antibiotics” (eds M. Salton and G . D. Shockman) pp. 249-60. Academic Press, New York Nishikawa, J. and Tori, K. (1981). J. Antibiotics 34, 1641, 1645 Nishikawa, J. and Tori, K. (1984). J. Med. Chem. 27, 1657

268

MICHAEL I PAGE

Nishikawa, J., Tori, K., Takasuka, M., Onoue, H. and Narisada, M. (1982). J. Antibiotics 35, 1724 Oakenfull. D. G. and Jencks, W. P. (1971). J. Am. Chem. SOC.93, 178 O’Callaghan, C. H. O., Kirby, S. M., Morris, A., Waller, E. R. and Duncombe, R. E. (1972). J. Bacreriol 110, 988 O’Connor, C. J. (1970). Quart. Rev. 24, 553 O’Connor, C. J. and Tan, A.-L. (1980). Austral. J. Chem. 33, 747 Page, M. I. (1973). Chem. SOC.Rev. 295 Page, M. I. (1976). Biochem. Biophys. Res. Commun. 72, 456 Page, M. I. (1977). Angew. Chem. (Int. Edn] 16, 449 Page. M. I. (1980a). Int. J. Biochem. 11, 331 Page, M. I. (l980b). In “Macromolecular Chemistry”, p. 397. Royal Society of Chemistry, London Page, M. I. (1981). Chem. Ind. 144 Page, M. I. (1984a). Accounts Chem. Res. 17, 144 Page, M. I. (1984b). In “The Chemistry of Enzyme Action” (ed. M. I . Page) p. 1. Elsevier, Amsterdam Page, M. I. and Jencks, W. P. (1972a). J. Am. Chem. SOC.94, 3263 Page, M. I. and Jencks. W. P. (1972b). J. Am. Chem. SOC.94, 8828 Page, M. I. and Proctor, P. (1984). J. Am. Chem. SOC.106, 3820 Page, M. I. and Crombie, D. A. (1984). In “Macromolecular Chemistry”, Vol3, p. 351. Royal Society of Chemistry, London Parker. C. W.. Shapiro. J., Kern. M. and Eisen, H. N. (1962). J . E.xp. Med. 115, 821 Parthasarathy, R. (1970). Actu Crystdogr. B26. 1283 Paschal, J. W., Dorman, D. E.. Srinivasan, P. R. and Lichter, R. L. (1978). J. Org. Chem. 43, 20 13 Paukstelis, J. V. and Kim, M. (1974). J. Org. Chem. 39, 1503 Pawelczyk, E.. Zajac, M. and Knitter, B. (1981). Pol. J. Pharmacol. Pharm. 33, 241 Petrongolo, C., Prescatori, E., Ranghino, G. and Scordamaglia, S . (1980). Chem. Phys. 45, 291 Pfaendler, H. R., Gosteli, J., Woodward, R. B. and Rihs, G . (1981). J. Am. Chem. SOC.103, 4526 Pitt, G. J . (1952). Acta Crysf.5, 770. Poortere, M. De., Marchand-Brynaert. J. and Ghosez, L. (1974). Angew. Chem. /Int. Edn] 13. 267 Pracejus. H., Kehlen, M.. Kehlen, H. and Matschiner, H. (1965). Tetrahedron 21, 2257 Pratt, R. F., Surh, Y. S. and Shaskus, J. J. (1983). J. Am. Chem. SOC.105, 1006 Proctor, P. and Page, M. I. (1979). Unpublished observations Proctor, P. and Page, M. I. (1984). J. Am. Chem. SOC.106, 3820 Proctor, P., Gensmantel, N. P. and Page, M. I. (1982). J . Chem. SOC.2 1185 Radom, L. and Riggs, N. V. (1980). Austral. J . Chem. 33, 249 Ree, B. and Martin, J. C. (1970). J. Am. Chem. SOC.92, 1660 Reinicke. B.. Blumel, P., Labischinski, H. and Giesbrecht, P. (1985). Arch. Microbiol. 141, 309 Richardson, F. S., Yeh, C.-Y., Troxell, T. C. and Boyd, D. B. (1977). Tetrahedron 33, 71 1 Romsted. L. S. (I 977). In “Micellisation, Solubilisation and Microemulsions” (ed. K. L. Mittal) Vol. 1, p. 509. Plenum Press, New York Schanck, A.. Coene. B., Van Meerssche, M. and Dereppe, J. M. (1979). Org. Magn. Reson. 12 337


269

Schanck. A., Coene, B., Dereppe, J. M. and Van Meersch, M. (1983). Bull. Soc. Chim. Belg. 92, 81 Schneider, C. H . and DeWeck, A. L. (1966). Helv. Chim. Acta 49, 1695, 1707 Schneider, C. H. and DeWeck, A. L. (1968). Biochim. Biophys. Acta 169, 27 Schowen, R. L., Jayaraman, H. and Kershner, L. (1966). J . Am. Chem. Soc. 88. 3373 Schrinner, E., Limbert, M. Penasse, L. and Lutz, A. (1980). J . Antimicrob. Chemother. 6, (Suppl. 1) 25 Schaartz, M. A. Granatek, A. P. and Buckwalter, F. H. (1962). J . Pharm. Sci. 51. 523 Schwartz, M. A. (1965). J . Pharm. Sci. 54, 472 Schwartz, M. A. and Delduce, A. J. (1969). J . Pharm. Sci.58, 1137 Schwartz, M. A. and Wu, G.-M. (1966). J . Pharm. Sci. 55, 550 Schwartz, M. A. (1968). J . Pharm. Sci. 57, 1209 Schwartz, M. A. (1969). J . Pharm. Sci. 58, 643 Schwartz, M. A. (1982). Biorg. Chem. 11, 4 Schwartz, M. A. and Pflug, G . R. (1967). J . Pharm. Sci. 56, 1459 Simon, G. L., Morin, R. B. and Dahl, L. F. (1972). J . Am. Chem. Soc. 94, 8557 Smith, C. R. and Yates, K. (1972). J. Am. Chem. SOC.94, 881 1 Spratt, B. G . (1975). Proc. Natl. Acad. Sci. USA 72, 2999 Spratt, B. G . (1977). Eur. J . Biochem. 72, 341 Spratt, B. G. (1980). Phil. Trans. Roy. Soc. 289, 27 Stackhouse, J., Bagchler, R. D. and Mislow, K. (1971). Tetrahedron Lett. 3437 Stewart, W. E. and Siddall, 111, T. H. (1970). Chem. Rev. 70 517 Stirling, C. J. M. (1979). Accounts Chem. Res. 12, 198 Strominger, J. L. (1967). Antibiorics 1, 706 Sweet, R. M. and Dahl, L. F. (1970). J . Am. Chem. Soc. 92, 5489 Sweet, R. M. (1973). In ‘Tephalosporins and Penicillins: Chemistry and Biology” (ed. E. H. Flynn,) p. 280. Academic Press, New York Sykes, R. B., Parker, W. L. and Wells, J. S. (1985). Trends in Anfibior. Res., I15 Taguchi, K. (1960). Chem. Pharm. Bull. (Tokyo) 8, 205 Takasuka, M., Nishikawa, J. and Tori, K. (1982). J . Antibiot. 35, 1729 Tipper, D. J. (1970). Int. J . Syst. Bacteriol. 26, 361 Tipper, D. J. (1985). Pharmac. Ther. 27, 1 Tipper, D. J. and Strominger, J. L. (1965). Proc. Nail. Acad. Sci. USA 54, 1133 Tomasz, A. (1979). Ann. Rev. Microbiol. 33, I I3 Tomasz, A. (1983). In “Handbook of Experimental Pharmacology (eds A. L. Demain and N. A. Solomon) Vol. 67/1, pp. 15-95. Springer-Verlag, Berlin Tomida, H. and Schwartz, M. A. (1983). J . Pharm. Sci.72, 331 Tori, K . , Nishikawa, J. and Takeuchi. Y. (1981). Tetrahedron Lett. 2793 Tsuji, A., Itatani, Y. and Yamana, T. (1977). J . Pharm. Sci. 66. 1004 Tsuji, A.. Yamana, T., Miyamoto, E. and Kiya, E. ( 1 975). J . Pharm. Pharmacol. 27, 580 Tsuji, A., Nakashima, E., Deguchi, Y., Nishide, K., Shimizu. T.. Horiuchi, S., Ishikawa, K. and Yamana, T. (1981). J. Pharm. Sci. 70, 1120 Tsuji, A., Miyamoto, E., Matsuda, M., Nishimura, K. and Yamana. T. (1982). J. Pharm. Sci. 71, 1313 Tsuji, A., Nakashima, E., Nishide, K., Deguchi. Y.. Hamano, S. and Yamana. T. (1983). Chem. Pharm. Bull. 31, 4057 Tutt, D. E. and Schwartz, M. A. (1971). J. Am. Chem. Soe. 93, 767 Vijayan. K., Anderson, B. F. and Hodgkin, D. C. (1973). J . Chem. Soc. Perkin Trans 1484

270

MICHAEL I. PAGE

Vishveshwara, S. and Rao, V. S. R. (1983). J. Mol. Struct. 92, 19 Wan, P., Modro, T. A. and Yates, K. (1980). Can. J. Chem. 58, 2423 Waxman, D. J. and Strominger, J. L. (1980). J. Biol. Chem. 255, 3964 Waxman, D. J. and Strominger, J. L. (1983). Ann. Rev. Biochem. 52, 825 Wei, C.-C., Borgese, J. and Weigele, M. (1983). Tetrahedron Lett. 1875 Weiss, A., Fallab, S. and Garlenmeyer, H. (1957). Helv. Chim. Acta 50, 576 Weiss, A. and Fallab, S. (1960). Helv. Chim. Acta 40, 61 Wheland, G. W. (1955). In “Resonance in Organic Chemistry”, pp. 367, 508. John Wiley and Sons, New York Williams, A., (1975). J. Chem. SOC.Perkin Trans 2, 947 Williams, A. (1976). J. Am. Chem. SOC. 98, 5645 Williamson, K. L. and Roberts, J. D. (1976). J. Am. Chem. Soc. 98, 5082 Wittig, G. and Steinhoff, G. (1964). Justus Liebigs Ann. Chem. 676, 21 Wojtkowski, P. W., Dolfini, J. E., Kocy, 0. and Cimarusti, C. M. (1975). J. Am. Chem. SOC.97, 5628 Wolfe, S. and Lee, W. S. (1968). J. Chem. SOC.Chem. Commun. 242 Wolfe, S., Godfrey, J. C., Holdrege, C. T. and Perron, Y. G. (1963). J. Am. Chem. Soc. 85, 643 Wolfe, S., Godfrey, J. C., Holdrege, C. T. and Perron, Y. G. (1968). Can. J. Chem. 46. 2549 Wolfe, S., Ducep, J.-B., Tin, K.-C. and Lee, S.-L. (1974). Can. J . Chem. 52, 3996 Woodward, R. B. (1949). In “The Chemistry of Penicillin” (eds H. T. Clarke, J. R. Johnson and R. Robinson) p. 443. Princeton University Press, Princeton, New Jersey Woodward, R. B. (1980). Phil. Trans. Roy. SOC. London. Ser. B 289, 239 Yamana, T. and Tsuji, A. (1976). J. Pharm. Sci. 65, 1563 Yamana, T., Tsuji, A., Kanayama, K. and Nakano, 0. (1974a). J. Antibiot. 27, 1000 Yamana, T., Tsuji, A. and Mizukami, Y. (1974b). Chem. Pharm. Bull. 22, 1186 Yamana, T., Tsuji, A., Miyamoto, E. and Kiya, E. (1975). J. Pharm. Pharmacol. 27, 283. Yamana, T., Tsuji, A., Kiya, E. and Miyamoto, E. (1977). J . Pharm. Sci. 66, 861 Yates, K. and Stevens, J. B. (1965). Can. J. Chem. 43, 529 Yatsimirski, A. K., Martinek, K. and Berezin, I. V. (1971). Tetrahedron 27, 2855 Yatsuhara, M., Sato, F., Kimura, T., Muranishi, S. and Sezaki, H. (1977). J , Pharm. Pharmac. 29, 638 Zugara, A. and Hidalgo, A. (1965). Rev. Real Acad. Cienc. E.ract. Fis. Nut. Madrid 59. 221

Free Radical Chain Processes in Aliphatic Systems involving an Electron Transfer Reaction G L E N A. R U S S E L L Department of Chemistry, Iowa State University, Ames, Iowa 50011 I Introduction 271 2 Free radical chain processes involving nucleophiles 274 Autoxidation of carbanions 274 Substitution by a mechanism involving radicals and radical anions 276 Processes leading to oxidative dimerization o r dehydrogenation of the anion 294 Processes involving reductive elimination 296 Processes involving hydrogen atom transfer from an anion 297 3 Free radical chain processes involving electron transfer between neutral substances 299 Reactions of 1,4-dihydropyridines, dialkylanilines, hydrazines, enamines and pyridines 299 Reactions of the tributyltin radical 303 Reactions of electron acceptor radicals with alkylmercury halides 306 4 Free radical chain reactions involving radical cations 308 Aliphatic substitution processes 308 Alkene dimerization and autoxidation 310 5 Concluding remarks 3 15 References 3 16

1

Introduction

There are only a limited number of elementary reactions that have been incorporated into free radical chain processes. The major propagation reactions of free radicals not involving electron transfer are: (1) addition to an unsaturated or coordinatively unsaturated system in an inter- or intra-molecular fashion; ADVANCES IN VOLUME 23

PHYSICAL

ORGANIC CHEMISTRY ISBN 0 I2 033523 9

271

Copyright 0 1987 Academic Press Inc. London All rights of reproduction in any form reserved

GLEN A

272

RUSSELL

( 2 ) elimination of the a, p. y or &-type; (3) S,2 atom or group transfer processes. Chain reactions not involving electron transfer involve appropriate combinations of (1)-(3) plus the required initiation and termination processes. Metal cations capable of existing in several oxidation states can participate in cyclic processes involving electron transfer wherein the metal ion cycles between two oxidation states. Oxidation-reduction couples such as Fe(lI)/Fe(lll), Cu(O)/Cu(I), Cu(I)/Cu(II) are effective in a number of such processes. In benzenediazonium ion chemistry, the Sandmeyer (Waters, 1942; Nonhebel and Waters, 1957), Meerwein (Koelsch and Boekelheide, 1944; Dickerman and Weiss, 1957; Dickerman et al., 1956, 1958; Kochi, 1957, 1967), and Cohen (Cohen et al., 1968) processes have been formulated as shown in Schemes 1-3 in homogeneous acetone solutions where CuCI, is reduced to CuCI. ArN,'

+ Cu"-Cu"+' + ArN,.-Ar A r + A - -ArA*+ Cu" ArA*- + Cunt'-ArA

+ N,

Scheme 1

ArN,

Ar. ArCH,CH(Ph).

- -+ -

' + CuCl

CuCl

+ CH,=CHPh

+ CuCI'

CuCl

+ N,

ArN,-Are ArCH,CH(Ph)-

+ ArCH,CH(Ph)+

CI

~

ArCH,CH(Ph)CI

Scheme 2

+ Cu(1)

-

CU(1l)

+ CONMe,

lip'

PhCONHMe

+ CH,O + H' Scheme 3

Peroxides, alkyl peresters or peracids can be forced to react in similar processes such as the Kharasch (Schemes 4, 5) (Kharasch et al., 1953;

273

FREE RADICAL CHAIN PROCESSES IN ALIPHATIC SYSTEMS

Kharasch and Sosnovsky, 1958) and Minisci (Schemes 6, 7) (Citterio ef al., 1977) substitution processes PhC(O)OOCMe,

+ Cu(1)

Me,CO + RH R- + Cu(I1)

-

Cu(I1) + PhCO;

-

+ Me,CO.

+ + + H+

Me,COH Re R + CU(I)

R + + SOH-ROS

Scheme 4

+

+

HORO. ROH + ROO.

ROOH Co(l1) RO. + ROOH

RO-(ROO.)

+

0-

ROH(RO0H) +

0

. + Co(II1)

+

+ Co(II1)

o+

-

+ ROOH(R0H)

+ Co(l1)

-0-

+ Hi

OOR(-OR)

Scheme 5

CU(1)

SO,'

--

+ s,o,2-

Cu(I1)

+ p-i-PrC,H,CH,

+ so,2 + SO,'

SO4'-

-

+ p-i-PrC,H,CH,*

H + + p-i-PrC,H,CH,-

+ cU(I) + H+

p-i-PrC,H,CH,- + cu(I1) -p-i-PrC,H,CH,+ p-i-PrC,H,CH, * + H,O-p-i-PrC,H,CH,OH

RC(0)OOH RCO,.

R.

H

+ PyH+ N

D

-

+ Fe(I1)

-

Scheme 6

Fe(lI1) + HO-

+ RCO,.

R. + C O ,

R

H

+ Fe(ll1)

N

3

eH

;

-

N

D

R

-

-

H

N

Scheme 7

O

R

+ Fe(l1)

+ H+

GLEN A. RUSSELL

274

The processes of Schemes 1-7 can be catalytic in the oxidation-reduction couple. The true mechanisms are often more complicated than the reaction schemes indicate, and the formation of various complexes involving the metal ion must be considered as well as the possible intervention of ligand transfer processes either within a cage or in noncage reactions. In Schemes 1-7 the “initiation” step is also one of the propagation steps and a clearly defined sequence of separate initiation, propagation and termination steps does not exist. The processes to be considered in this review will be restricted to those where the propagation and initiation reactions are clearly different and where one member of any oxidation-reduction couple has a low persistency and can be considered to be a reactive intermediate. Among the processes to be considered which can be incorporated into chain reactions are: (i) electron transfer between a radical and an anion (ii) electron transfer from a radical anion to a neutral reducible substance (iii) electron transfer from an easily oxidized neutral radical to a neutral reducible substance (iv) electron transfer to a reducible radical or radical cation from a neutral reducible substance 2

Free radical chain processes involving nucleophiles

AUTOXJDATJON OF C A R B A N I O N S

The autoxidations of certain carbanions, such as fluorenide anion, have been demonstrated to proceed via a chain mechanism involving electron transfer (Russell, 1953; Russell et af., 1965. 1968) as shown in Scheme 8. In Me,SO Re + O,-ROOR O O + R:--ROORROO- + Me,SORO- + Me,SO, Ar,CHOOH + B- -Ar,C----O + BH + OHROOH + RRO. + Re + HOScheme 8

+

-

the intermediate hydroperoxide is cleanly converted into the alkoxide which may be further oxidized via the dianion to the ketone. Carbanion autoxidation can be initiated by electron acceptors such as nitroaromatics as shown in ( I ) and (2). The rates of electron transfer from fluorenide anions to R:- -I- ArNO, ArNO,’ + 0 ,

+ -

R- ArNO,’ ArNO, + 0,;


275

nitroaromatics in the absence of oxygen (Russell et al., 1964) parallels the catalysed autoxidation rates (Russell et al., 1962a, 1968). With Ph,CH- and Ph,C-, the free radical chain is apparently rapidly initiated by electron transfer from the carbanion to molecular oxygen, and the rate of autoxidation of the parent hydrocarbon is equal to the rate of its ionization with Me,COK in Me,S0(80%)-Me3COH(20%) solution (Russell and Bemis, 1965). With fluorene the rate of oxidation becomes equal to the rate of ionization only in the presence of large amounts of the excellent electron acceptor p-CF,C,H,NO, (Russell and Weiner, 1969). The ease of autoxidation of many carbanions follows the sequence, nC(CH,), > 7cCHCH; > nCH, (n = RCO, RO,C, O,N, ArSO,) and reflects the ease of electron transfer from the carbanion to 0, or ROO. (Russell et al., 1962b). The autoxidation of Grignard reagents or dialkylmercurials proceeds by a free radical chain mechanism involving the attack of ROO. upon RMgX (Walling and Buckler, 1955; Lamb et al., 1966) or R,Hg (Razuvaev et al., 1960; Aleksandrov et al., 1964). These reactions may well involve electron transfer or at least a transition state for the formal S,2 substitution reaction resembling [ I ] [ROO-

dY

Re]

[I1 The anion of 2-nitropropane in EtOH/EtOLi will undergo a free radical chain autoxidation (Russell, 1953). However, the anion under these conditions does not undergo an electron transfer reaction with molecular 0,. The reaction can be initiated by the presence of a nitroaromatic or by the unionized 2-nitropropane according to (1) and (2) (Russell et al., 1965). The in'itiation step can be accelerated by photolysis whereby the electron acceptor is excited to the n* state. The chain reaction of Me,C=NO; with 0, is autocatalytic suggesting the formation of O,NCMe,OOH (Scheme 8). However, the major reaction pathway appears to follow Scheme 9 (Russell, O,NCMe,.

+ 0,

O,NCMe,OO.

-

+ Me,C=NO;

O,NCMe,OOCMe,NO;

O,NCMe,O.

O,NCMe,OO.

-c

0, NCMe,OO

-+

+ Me,C=NO;

NO;

0, NC Me, OOCM e2N 0,T O,NCMe,OOCMe,. O,NCMe,O. Me,CO

--

O,NCMe,. NO;

Scheme 9

+ 0,NCMe, .

+

+ O,NCMe,O + Me,CO

276

GLEN A. RUSSELL

1967) and involves the addition of the peroxy radical to the nitronate anion with the ultimate production of Me,C==O and NO, in nearly quantitative yield. As shown in Scheme 9, an intermediate radical anion of an aliphatic nitro compound is formed which then decomposes to yield NO;, Me,C=O and an alkoxy radical which can continue the chain reaction. Scheme 9 embodies the basic steps involved in many free radical chain processes. The free radical chain propagates not only by an electron transfer step but also by the addition of a radical to an anion to yield a radical anion which subsequently fragments. These steps are of importance in the so-called (Kim and Bunnett, 1970) S,,1 sequence of Scheme 10.

-

+ N--RNT + RX RX' + RN RXTR. + X -

R. RN'

(3)

Scheme 10

SUBSTITUTION BY A MECHANISM INVOLVING RADICALS A N D RADICAL ANIONS

The S,,I process of Scheme 10' was originally recognized in the reactions of Me,C=NO; with O,NCMe,CI or p-nitrobenzyl halides (Kornblum et al., 1966; Russell and Danen, 1966, 1968). In the electron transfer step of Scheme 10, the electron is transferred from an easily oxidized radical anion (RN-) to an easily reduced substrate (RX). The importance of the nitro group in these substitution reactions must be connected with the low energy of its LUMO which leads to stability of R N r whenever R or N contains a nitro group. In Table I are collected a number of nitro-containing substrates (RX) which will undergo the S,,I reaction with the anions shown. The nitro substituent in R can be at the carbon bonded to the leaving group X as in O,NCR,X or in a vinylogous position such as p-O,NC,H,CR,X. Substrates such as p-O,NC,H,C(O)CR,X are also reactive in the S,, 1 process, perhaps because of intramolecular electron transfer between the nitro group and the carbonyl group giving rise to the ketyl radical anion which readily fragments. Radicals such as Me,C- or

' The acronym S,,I connotates a free radical chain Substitution reaction in which the new bond is formed by attack of a Radical (Re) upon a Nucleophile ( N - ) and in which the bond to the nucleofuge (X) is broken in a unimolecular process. When allylic rearrangement is involved, the process is termed S,, 1'.

TABLEI Examples of aliphatic SRNlsubstitutions with nitro-containing substrates (RX) (RX

+ N--RNhv

+ X-) R

X

N-

Reference

Me,CNO,

C1, Br, I, ArSO,, Me,C=NO,-, c-C,H,,=NO;, PhS, p-CIC,N,S, (EtO,C),CR-, (EtO),PO-, P-O,NC~H.+S, (EtO),PS-, N;, ArSO;, ArSo-0,NC6H,S, PhSO, (Ar = 2-pyridy1, 4-pyridy1, p-CIC6H,S0, SCN, pyrimidin-2-yl, 4-5-dihydroNO2 1,3-thiazol-2-yl, 1-methylimidazol-2-yl, 1-3benzothiazol-2-yl, p-CIC,H, p-O,NC,H,, o-O2NC6H4

R~R~CNO, R', R2 = Me; Me, Et; -(CH2)5-

CI, Br

RC(CN)CO,Et- (R = i-Pr, PhCH,), (EtO,C),CR- (R = Et, n-Bu, PhCH,), EtO,CC(R)COMe(R = Et, n-Bu, PhCH,), (MeCO),CR- (R = Me, n-Bu) 0&OM.

0&02Et

Russell and Danen, 1966, 1968; Russell et al., 1971, 1982d; Russell and Hershberger, 1980a; Kornblum and Boyd, 1970; Kornblum et al., 1971, 1973, 1974; Kornblum, 1975; Zeilstra and Engberts, 1973; Bowman and Richardson, 1980, I98 1 ; Bowman and Symons, 1983; Bowman et al., 1984; Al-Khalil et al., 1986 Ono et al., 1983

0-

0 Y 2 E t

TABLE1 (continued)

X

R Me,CNO,

SAr (Ar = 1 -methylimidazol-

2-yl, pyrimidin2-yl, 2-pyridy1, 1,3-benzothiazol2-yl, 4,5-dihydroI ,3-thiazol-2-yl)

N-

Reference

R S - (R = p-chlorophenyl, benzyl, L-cysteine), Me,C==NO;

Bowman et al., 1984

Beugelmans et al., 1983

Beugelmans et al., 1983

MeC(R)NO, R = CH,OTHPO CH,CH,CO,Me

CI

MeO,CCH,CH,C(CH, CH,OTHP)NO, NCCH,CH,C(CH(CH,) OTHP)NO, c-C,H,ONO,

CI

(EtO,C),CH-, (EtO,C),CEt-, EtO,CC(CH,)COCH , CH,CH,C(CH,OCH,)=NO;, CH,C(CH,CH,COCH,)=NO 2, c - C ~ Ho=NO, , RCOCHC0,Et -, R'COCHC0R'-,

Crozet and Vanelle, 1985; Russell et al., 1981b

I

Et

THP

=

tetrahydropyranyl

(EtO,C),CEt Me,C(Z) Z = C O R , CO,R, CN, p-0,NC6H4N=N R'R2CCN R', R2 = Me; Me, Et; Me, Me(CH,),; -(cH2)4-> -(CH2)11Me,CN, R'COCHR~, R'O,CCHR~, PhCHR' R' = Ph, Me, Et, i-Pr R' = H, Me, Et, n-Pr p-R 'C6H,C( R2)Me R' = H, CN, PhSO,, PhCO, 3,5-(CF3)2 R2 = Me, Et Me,C R,CMe, R, = n-C,F,,, n-C,F,, p-YC,H,CMe, Y = PhSO,, CN p-YC,H,C(R)MeCMe, Y = PhSO,, CN, PhCO, R = Me, Et 3,5-(CF3),C6H,CMe(R') CMe(R2) R' = Et, R2 = Me R' = Me, R Z = Et PhCOCMe,

Me,C==NO; Me,C==NO;

Russell et al., 1971 Russell et al., 1971; Kornblum and Boyd, 1970

R~R~C=NO; R', R2 = H; Me; Me, H; Et, H; Me(CH,),, H; -(cH2)4-> -(CH*)G 4CH2)ll-r PhS-, p-ClC6H4SPhS-, p-ClC,H,Sp-ClC6H4S0L1
c(No,)-c(No,)~

I

RS-C-C(NO,T),

/

I

-

-

NO;

+

+ ;c(No;)-c(No,< )k-C(NO,): + NO,

-

RSSR

-

Scheme 23

PROCESSES I N V O L V I N G H Y D R O G E N ATOM T R A N S F E R FROM A N A N I O N

A variety of free radical chain processes are known in which reactions (37)-(39) are coupled (Scheme 24). Among the negatively charged hydrogen R- + HY--RH Y: + RX-Y R-X; R.

-

+ Y; + RX; +X-

Scheme 24

donors (HY-) recognized to react in this manner are BH; (Groves and Ma, 1974), AIH; (Hatem and Waegell, 1973; Chung and Chung, 1979; Chung, 1980; Singh et al., 1980, 1981; Singh and Khanna, 1983; Ashby et al., 1984), cyclopentadienyl-V(CO),H - (Kinney et al., 1978). CH,O- (Simig and Lempert, 1961; Simigetal., 1977, 1978), CH,S- (Kornblum et al., 1978a, 1979a,b), and PhSiMe,HF- (Yang and Tanner, 1986). Table 3 lists some typical aliphatic substrates which participate in the chain process of Scheme 24. The chain reactions involving LiAlH, in T H F are presumed to occur with a variety of vinyl, cyclopropyl and bridgehead halides as well as aryl

TABLE3 Free radical chain reactions involving Scheme 24 +XU-) (RX + HY--RH

R Ph,C(CONMe,)

X Br

HYCH,O-

Reference

I NO2 SMe

CH,OCH,S-, CH,SCH,SCH,S-, CH,SCH,S-

Simig and Lempert, 1961; Simig al., 1977, 1978 Boyle and Bunnett, 1974 Kornblum et a / . , 1978a, 1979a,b Kornblum et a/.. 1979a,b

Br Br CI, Br

AIH 4 BH 4 AIH 4

Chung, 1980 Groves and Ma, 1974 Singh er al., 1981

Br, CI HgCl F I

HCpV(C0); AIH PhSiMe,H- - - F AIH;, AIH,

Kinney et al., 1978 Singh and Khanna, 1983 Tanner et al.. 1986a Ashby et al., 1984

el

Me,CCH, 3 -Alkyl, 3 -benzylic p-XC,H,CMe, X = PhSO,, CN, PhCO ( E ) and ( Z ) PhCH=CH 7.7-Norcarane Benzyl, diphenylmethyl. 9-fluoreny1, anthracenecarbinyl Alkyl, acyl Alkyl PhCOCH, 2-Octy1, 2-hepten-6-yl, 1-hexen-5-yl; 2,2-dimethyl- I -hexen-5-yl


299

halides (Barltrop and Bradbury, 1973; Chung and Chung, 1979; Singh et al., 1980; Chung and Filmore, 1983). Free radical chain reactions initiated by AIH, or A.!HS are recognized to occur in the reduction of alkyl iodides such as 2-iodooctane, 6-iodo- 1-heptene, 6-iodo- 1-hexene, 2,2-dimethylI-iodo-5-hexene (Ashby et al., 1984). Similar reactions of CH,O- are also recognized in the reductive substitution reaction of aromatic systems, particularly p-N02C6H41 or p-NO,C,H,N=NOMe, p-NO,C,H,N; (Detar and Turetzky, 1955, 1956; Bunnett and Takayama, 1968a,b; Boyle and Bunnett, 1974). When the reduced product (RH) of Scheme 24 can be deprotonated by the basic HY- to give R - , the S,,I chain reaction between R - and RX (Scheme 10) may occur as exemplified by (40) (Zorin et al., 1983). Br NaSEt

2

3 Free radical chain processes involving electron transfer -etween neutral substances R E A C TI 0 N S 0 F I , 4 - D I H Y D R 0 P Y R I D IN ES, D I A L K Y L A N I L I N ES, H Y D R A Z I N ES, E N A M I N E S A N D , P Y R I D I N E S

Easily oxidized neutral radicals can be one-electron donors to easily reduced neutral molecules as in Scheme 20. The reverse process involving neutral acceptor radicals and neutral oxidizable substrates is of course possible but is not a commonly recognized chain propagation reaction in free radical

R

CH,Ph [81

[7] (R = Me, H)

processes. 1 ,4-Dihydropyridines, such as [7] and [8], participate in a variety of chain processes by virtue of the ease with which reactions (41) and (42) occur where nH2 is a dihydropyridine and A is a neutral acceptor molecule. R. + rrH,-RH TH + A-rcH'

+ TH + A'

300

GLEN A. RUSSELL

When AT can regenerate R. by a fragmentation process, a chain reaction will ensue. Thus [7] is oxidized to the pyridinium cation [7]+ in chain reactions by electron acceptor molecules such as MeC(O)OOC(O)Me, PhC(0)OOC(O)Ph, t-BuO,C(CH,),, t-BuOOBu-t, BrCCl,, I,, HOOH or S , 0 , - 2 (Huyser et al., 1964, 1972; Huyser and Kahl, 1970; Huyser and Harmony, 1974). Among the oxidants which are effective in the conversion of [8] to the cation are tetramethylthiuram disulfide (Huyser and Harmony, 1974), alkylmercury acetates or alkylthallium dichloride (Kurosawa et al., 1980, 198l), nitroalkanes such as R,C(NO,)Br, R,C(NO,)CN, R,C(NO,)CO,R, R,C(NO,)COAr, R,C(NO,)SO,Ar (Kill and Widdowson, 1976; Ono et al., 1980, 1985), and PhCOCH,X with X = F, C1, Br (Tanner et al., 1986). Electron transfer to these reagents forms, after fragmentation, the species MeCO,. (or Me-), PhCO,., CCl,., I., Me,N., alkyl., R,C(NO,)., or PhCOCH,. which serve as R. in Scheme 25. In the case of RHgCl or BrCCI,, the electron transfer from [7]. or [8]. is probably dissociative while for the substituted nitroalkanes and probably for the peresters the radical anion (RXT) is a true intermediate. R- + [7]or [8]-[7]or [8]* + RH or [8]+ + RX' or R. + X [7]. or [8]- + RX-[7]' RXT-

+

R. X Scheme 25

Reduction of PhCOCH,Br to PhCOCH, by NADH occurs by the mechanism of Scheme 25 (Tanner et al., 1986). The reaction is not catalysed by horse liver alcohol dehydrogenase, and the free radical chain dehydrogenations of 1,4-dihydropyridines seem not to be involved in enzymatic reactions involving NADH as a hydrogen donor. The formal hydride transfer between NAD+ analogues has been interpreted as a multistep but nonchain process involving electron and proton transfers (Ohno ef al., 1981) or alternatively as a one-step hydride transfer with an appreciable primary isotope effect (Ostovic et al., 1983). Although not as yet recognized, a free radical chain mechanism should be possible in certain cases where [8]*transfers an electron to an easily reducible NAD' analogue such as N-methylacridinium ion. The chain could continue by hydrogen atom abstraction from [8] by the N-methylacridinyl radical. The reaction of substituted aminomethyl radicals with benzoyl, acetyl or cyclohexanesulfonylacetyl peroxides involves at least partially the free radical process of Scheme 26 (Horner and Schwenk, 1944; Horner and Betzel, 1953; Horner and Anders, 1962; Walling and Indictor, 1958; Hrabak and Lokaj, 1970). The same products are also formed by the ionic Polonovskitype reaction involving the intermediate Phy(Me),O,CPh+ PhCO, which decomposes to PhN(Me)CH,O,CPh via PhN(Me)(O,CPh)CH; .


PhN(Me)CH,.

+ PhC(O)OOC(O)Ph

-

+

301

+

PhA(CH,)=CH, PhCO,. PhCO; PhCO,-Ph. + CO, Ph.(PhCO,-) + PhNMe, PhH(PhC0,H) + PhN(CH,)CH,Phq(CH,)CH, PhCO; PhN(CH3)CH,0,CPh Scheme 26

+ -

Oxidation of phenylhydrazine to phenyldiimide by BrCCI,, halogens or S,O, - appears to involve similar processes (Scheme 27).

-

PhNHNH, + CI,C.-PhNNH, PhN=NHi PhNNH, + BrCCl,

+ HCCI, + Br- + CI,C.

Scheme 21

The electron donor radicals in Schemes 26 and 27 are generated by hydrogen abstraction reactions by radicals which themselves are either or have little donor ability such as Me. acceptors (PhCO,., Cl,C-, I., SO,:) or Ph.. Another route to a donor radical is the addition of a radical to an unsaturated system such as an enamine or ynamine. Thus, the addition of CF,., *CF,Cl, CF,Br or CF,CF,Br to an enamine converts an acceptor radical into a donor radical (e.g. R,NCHC(CF,)R,) and the free radical process of Scheme 28 occurs (Cantacuzene and Dorme, 1975; Rico et al., 1981, 1983).

ON3 -

Rf'+ +

[9] RfX

-

QN3 [91

[9]++ R,. + X-

kf

Scheme 28

Addition of p-nitro-a,a-dimethylbenzyl radical to a variety of coordinatively unsaturated primary, secondary or tertiary amines or ammonia produces the radical zwitterion [lo] (Kornblum and Stuchal, 1970). The radical zwitterion will transfer an electron to p-nitrocumyl chloride in the S,,1 process.

GLEN A. RUSSELL

302

Addition of alkyl radicals to the pyridine (43) or pyridinium ring (44) will also lead to our easily oxidized pyridinyl radical. Radical substitution in pyridinium ions occurs readily in a variety of reactions involving H,O,,

NH,OH, t-BuOOH or PhC(O)OOC(O)Ph and metal ions such as Ti(II1) or Fe(II), e.g. Scheme 7 (Minisci et al., 1985). A reaction (Scheme 29) can be observed in the absence of metal ion where the solvent (SH) is a primary alcohol and S. an a-hydroxyalkyl radical (Minisci et al., 1985).

I H

I

H

- + + -

[ I 11 + PhC(O)OOC(O)Ph PhCO;

Ph.(PhCO;)

Ph-

SH

[ I I]'

I

H

+ PhCOI + PhCO,.

CO,

PhH(PhC0,H)

Scheme 29

+ S.


303

Alkylmercury halides or carboxylates undergo a free radical aromatic substitution reaction with pyridine, pyridinium ions or dialkylanilines. Addition of R- to these substrates produces the easily oxidized radicals

NMe,

H

1121

~ 3 1

[I I]-[I31 which undergo dissociative electron transfer with RHgX to regenerate R. as shown in Scheme 30 (Russell et al., 1985). Since the stability of R- formed in the dissociative electron transfer step controls the rate of this

[I21

+ RHgCl

-

[12It

+ R. + Hgo + CI-

Scheme 30

process, it follows that the overall rate of reaction of pyridine with RHgCl increases according to the order R = n-Bu < i-Pr < t-Bu. In reaction with N,N,N',N'-tetramethyl-p-phenylenediamine, it is also observed that PhCH,HgCI is more reactive than n-BuHgC1. The preferred position of attack of an alkyl radical upon pyridine is at the a-position to give [12] after tautomerization. The ratio of o/p attack by R increases with the donor ability of R- according to the order R = n-Bu < i-Pr < t-Bu. Attack of R. upon pyridinium ion occurs with a slight preference at the p-position to yield [ 1 11; again the p/o ratio is a function of the donor ability of R. and increases from R = n-Bu to i-Pr to t-Bu. R E A C T I O N S O F THE T R I B U T Y L T I N R A D I C A L

The tributyltin radical is easily oxidized to the cation. Therefore, it seems reasonable that Bu,Sn- should transfer an electron to an easily reduced substrate such as R,CNO,, ArCH,I or RHgX as in (45). This process has Bu,Sn*

+ RX

-

Bu3Sn+ + X -

+ R-

(45)

GLEN A. RUSSELL

304

been reported to occur in the reaction of nitroalkanes (Tanner et al., 1981) and benzylic iodides (Blackburn and Tanner, 1980). In reaction of Bu,Snwith an aliphatic nitro group, an alternative chain reaction (46) not involving electron transfer must be considered (On0 er al., 1985). The observation R-NO,

+ Bu,Sn.-

R-N-0-SnBu,

I

-+ R.

Bu,SnONO

(46)

-0

that Me,C(SO,C,H,CH,-p)NO, undergoes reaction with [8] according to Scheme 30 but fails to undergo any reaction with Bu,SnH has been interpreted as evidence that (46) is often the preferred route of reaction for Bu,Sn-. The reaction of Bu,SnH with alkylmercury chlorides does appear to involve the electron transfer of (45). When Bu,SnH is utilized in the Giese reaction (Giese and Horler, 1985) with CH,=CH(Cl)CN (Scheme 31), it is

+

-+

R* CH,=C(CI)CN RCH,C(Cl)CN RCH,C(CI)CN Bu,SnH RCH,CH(CI)CN Bu,SnBu3Sn+ CI- + Hgo RBu,Sn- RHgCl-

+

+

+ +

Scheme 31

observed that in a direct competition, t-BuHgC1 is more than 100 times as reactive as n-BuHgC1 (Russell, 1986). In another reaction involving Bu,Sn., it is observed that the reaction of (E)-PhCH=CHSnBu, with RHgX forms (a-PhCH=CHR with a relative reactivity of R = t-Bu : c-C,H, : n-Bu of 1 .O : 0.01 I : 0.001. The substitution process of Scheme 32 occurs in the

+

--+

R. PhCH=CHSnBu, PhCHCH(R)SnBu, PhCHCH(R)SnBu, PhCH=CHR + Bu,SnBu,Sn. RHgClBu,Sn+ CI- + Hgo R.

+

+

Scheme 32

reaction of vinyl or acetylenic tin, compounds with RHgCl (Russell et al., 1984; Russell and Ngoviwatchai, 1985) and involves the sequence of addition-elimination-electron transfer. Figure 3 illustrates the relative reactivity observed in the electron transfer reaction of Bu,Sn. with a 1 : I mixture of t-BuHgC1 and n-BuHgC1 (> 100 : 1) as well as the unselective reaction observed in the BH, reduction ( - 1 : 1) which involves Hg(1) intermediates (i.e. Scheme 14). Data for the reaction of LiAlH, in T H F with RHgCl is included in Fig. 3 which indicates that -80% of the reaction occurs by the AIH,; electron transfer mechanism (Scheme 24) and 20% by the mechanism involving RHgH.


R,HgCI

R,CH,CH(CI)CN

+ MH + CH,=C(CI)CN R,HgCI

+

305

+ MCI

R,CH,CH(CI)CN

1.0

2.0

3.0

It-B~HgCI]~/[Alkene], FIG.3 Relative reactivities (k,/k,) of t-BuHgCl(H) and n-BuHgCl(H) towards RCH,C(CI)CN or RCH,CHCO,Et in competitive reactions with CH,=C(CI)CN and CH,=CHCO,Et at 25°C. (Data are from Russell, 1986.) Reactions in the presence of BH; involve hydrogen atom abstraction from RHgH, while reactions in the presence of Bu,SnH involve electron transfer from Bu,Sn. to RHgCl

A reduction mechanism analogous to Scheme 24 has been observed with the easily oxidized Ph,Sn.. The reaction of Ph,SnH with PhCOCH,F (Tanner et al., 1985b), PhCOCH,CI or PhCOCH,Br (Tanner et al., 1986b) proceeds via the loss of F-, C1- or Br- from PhCOCH,XT to yield PhCOCH,. which reacts with Ph,SnH to form PhCOCH, and Ph,Sn-. The chain is continued by electron transfer from Ph,Sn- to PhCOCH,X. Reduction of phenyl cyclopropyl ketone by Ph,SnH but not by (n-Bu),SnH also proceeds via electron transfer from the tin radical to form the ketyl which undergoes p-elimination and hydrogen atom abstraction to yield n-butyrophenone in methanol solution (Tanner ef al., 1985b).

i

306

GLEN A RUSSELL

REACTIONS OF ELECTRON ACCEPTOR R A D I C A L S W I T H ALKYLMERCURY HALIDES

Alkylmercurials can serve as electron donors to an electron acceptor radical. It is, however, difficult to distinguish the electron transfer (47) from a S,2 reaction occurring in one step (48). Electron accepting radicals such as A.

-c

+ RHgX

[A:- +HgX + R-I-AHgX AHgX

+ Re

+ R*

(47) (48)

PhS., PhSe., PhTe-, PhSO,. or t-BuCH,CHE with E = (EtO),PO, PhSO, or p-NO,C,H, participate in such reactions (Russell and Tashtoush, 1983; Russell et al., 1986a). The relative reactivities of t-BuHgCI, i-PrHgCI, and nBuHgCl are shown in Table 4 (Russell, 1986). It is also reported that PhCO,. reacts with R,Hg with a rate increasing according to the sequence R = Me < I"-alkyl < 2O-alkyl < 3O-alkyl (Nugent and Kochi, 1977). The reactions of Scheme 33 occur with A- = PhSO,., PhSO., PhS., I- (Russell et

-

R- + RCH=CHA Rr-
=.i

= adamantylideneadamantane

Scheme 44

tures the dioxetane radical cation [22r has been observed by esr spectroscopy (Nelsen et al., 1986). Unsymmetrically substituted alkenes produce the dioxetanes nonstereospecifically as expected for a mechanism involving [ 2 l r as an intermediate (Ando et al., 1982). Reactions involving singlet oxygen can be excluded since '0, fails to form the dioxetane from olefins with unprotected alkyl groups. Thus, isopropylideneadamantane reacts with 30,by the chain mechanism of Scheme 44 to form the dioxetane but reacts with '0, to yield only the allylic hydroperoxide (Nelsen et al., 1984; Nelsen and Teasley, 1986). Under proper conditions, chain reactions involving radical cations should be important in cis-trans isomerizations of alkenes and 1,2-disubstituted cyclopropanes, particularly with aryl substituents (Roth and Schilling, 1979, 1980a). There is abundant evidence that electron transfer is involved in the cis-trans isomerizations of 1,2-diarylcyclopropanes photosensitized by naphthalenes and cyano-substituted aromatics (Hixson et al.,1974; Hixson, 1979; Wong and Arnold, 1979) and in the isomerization of 1,1,2,2-tetraphenylcyclopropane to I , 1,3,3-tetraphenylpropene(Arnold and Humphreys, 1979; Wayner and Arnold, 1985). However, on the basis of CIDNP signals observed in the presence of chloranil, it appears that the isomerization between the radical cations of cis- and trans-l,2-diphenylcyclopropane occurs slowly (or not at all) in competition with the back electron transfer from the chloranil radical anion (Roth and Schilling, 1980b, 1981). Processes initiated by electron transfer to radical cations such as Ar,N*, or more powerful oxidizing agents which can act in an irreversible manner, would have a better chance to occur by a chain process in which the rearranged radical cation continues the chain by electron transfer with the unreacted alkene or cyclopropane.


5

31 5

Concluding remarks

The occurrence ol' electron transfer reactions greatly expands the scope of free radical chain processes and provides a crossing of the boundary between polar and homolytic processes. The formation of a radical ion from a neutral substrate or a radical from an ion can be considered to be a type of activation (Chanon, 1982, 1985). Electron transfer between radical ions and easily oxidized or reduced substrates is a common process. Somewhat less common in the absence of metal ions are electron transfers between neutral radicals and easily oxidized or reduced diamagnetic substances. The mechanistic possibilities for free radical chain processes are greatly expanded by the recognition that there is a bridge between radical ion chemistry and the chemistry of separated radicals and ions since in appropriate cases radical ions can dissociate to form a radical and an ion, or conversely, radicals may add to ions to form radical ions. It is indeed this bridging of the chemistry of radical ions and of free radicals which is the basis of numerous thermal or photostimulated substitution processes in organic, organometallic and inorganic chemistry of which the SRN1process is perhaps the most notable example. The occurrence of free radical chain processes involving electron transfer is, of course, most likely to occur when easily oxidized or reduced substrates are involved, with radical anion formation being favoured by strongly basic or reductive conditions and radical cations by acidic or oxidative environments. The experimental proof of a free radical chain mechanism usually requires relatively simple kinetic analysis although questions inevitably remain concerning the exact timing and nature of some of the individual steps which contribute to the overall chain reaction. Electron transfer can, of course, also be involved in multistep but nonchain processes which interconvert diamagnetic reactants and products (Russell rt al., 1964; Eberson, 1982). In general, in these cases one expects a spectrum of substitution mechanisms with the radical chain processes representing one extreme and reaction proceeding via transition states involving an electron shift process (Pross, 1985) as the other extreme. In between these extremes are multistep but nonchain processes involving intermediates, either free or in a cage, resulting from electron transfer processes. However, reaction by the radical chain mechanism is not in all cases simply an alternative to a one-step polar process since the product and the selectivity observed can be quite different in the two processes. This is particularly well illustrated in the reactions of Me,C=NO; which undergo alkylation at oxygen in typical SN2processes but which undergo carbon alkylation in S,, substitution reactions.

31 6

GLEN A

RUSSELL

References

Alder. R. W. (1980). J . Chem. Soc., Chem. Cownun. 1184 Alder. R. W., Casson, A. and Sessions. R. B. (1979). J . Am. Chem. Soc. 101, 3652 Aleksandrov. Yu. A., Druzhkov, 0. N. Zhil'tzov, S. F. and Razuvaev, G. A. (1964). Dokl. Akad. SSSR 157. 1395 Al-Khalil, S. I., Bowman, W. R. and Symons, M. C. R. (1 986). J . Chem. SOC.,Perkin Trans 1 5 5 5 Ando, W., Kake, Y. and Takata, R. (1982). J . Am. Chem. SOC.104, 7314 Arnold, D. R. and Humphreys, D. W. R. (1979). J . Am. Chem. SOC.101, 2743 Ashby. E. C. and Argyropolous, J. N. (1985). J. Org. Chem. 50, 3274 Ashby, E. C., DePriest, R. N.. Goel, A. B., Wenderoth. B. and Pham. T. N. (1984). J . Org. Chem. 49, 3545 Ashby. E. C.. Park, W. S.. Goel, A. B. and Su, W.-Y. (1985a). J . Org. Chem. 50,5184 Ashby, E. C.. Su. W.-Y. and Dham. T. N. (1985b). Organometallics 4, 1493 Bank, S. and Noyd. D. A. (1973). J . Am. Chem. SOC.95, 8203 Barker, S. D. and Norris, R. K. (1979). Tetrahedron Lett. 1973 Barltrop, J. A. and Bradbury, D. (1973). J . Am. Chem. SOC.95, 5085 Barton, D. H. R., LeClerc, G., Magnus, P. D. and Menzies, 1. D. (1972). J . Chem. Soc., Chem. Commun. 447 Barton, D. H. R., Hayes, R. K., Magnus, P. D. and Menzies, I . D. (1974). J . Chem. Soc.. Chem. Commun. 51 1 Barton, D. H. R.. Hayes, R. K., LeClerc, G., Magnus, P. D. and Menzies, I. D. (1975). J. Chem. SOC.,Perkin Trans. I 2 0 5 5 Bauld, N. L., Bellville, D. J. and Pabon, R. A. (1983a) J . Am. Chem. SOC.105, 5158 Bauld, N. L., Bellville, D . J., Pabon. R. A., Chelsky, R. and Green, G. J. (1983b). J . Am. Chrm. Soc. 105. 2378 Bellville. D. J. and Bauld, N. L. (1982) J . Am. Chem. SOC.104, 2655 Bellville. D. J., Wirth, D. D. and Bauld, N. L. (1981) J. Am. Chem. SOC.103, 718 Bellville, D. J., Bauld. N. L.. Pabon, R. A. and Gardner, S. (1983) J . Am. Chem. SOC. 105, 3584 Bethell. D.. Handoo. K. L., Fairhurst, S. A. and Sutcliffe, L. H. (1979). J . Chem. SOC.,Perkin Trans 2 707 Beugelmans, R., Lechevallier, A. and Rosseau, H. (1983). Tetrahedron Lett. 1187 Beugelmans, R., Lechevallier, A. and Rosseau. H. (1984). Tetrahedron Lett. 2347 Blackburn, E. V. and Tanner, D. D. (1980). J . Am. Chem. SOC.102, 692 Boiko. V. N.. Shchupak. G. N. and Yagupol'skii, L. M. (1977). J . Org. Chem. U S S R , [Engl. Trans.J 13. 1972 Bordwell, F. G. and Clemens. A. H. (1982). J . Org. Chem. 47, 2510 Borsub, N. and Kutal, C. (1984). J . Am. Chem. SOC.106, 4826 Bowman, W. R. and Richardson, G. D. (1980). J . Chem. SOC.,Perkin Trans I 1407 Bowman, W. R. and Richardson, G. D. (1981). Tetrahedron Lett. 1551 Bowman, W. R. and Symons, M. C. R. (1983). J . Chem. SOC.,Perkin Trans 2 25 Bowman. W. R.. Rakshit, D. and Valmas, M. D. (1984). J . Chem. SOC.,Perkin Trans I2327 Boyle, W. J. and Bunnett, J. F. (1974). J . Am. Chem. SOC.96, 1418 Bruce, M. I., Kehoe, D. C.. Matisons, J. G.. Nicholson, B. K., Rieger, P. H. and Williams, M. L. (1982). J . Chem. SOC.Chem. Commun. 443 Brunet, J.-J., Sidot, C. and Caubere, P. (1983). J . Org. Chem. 48, 1166 Bunnett. J . F. and Takayama, H. (1968a). J . Org. Chem. 33, 1924 Bunnett, J. F. and Takayama, H. (1968b). J . Am. Chem. SOC.90, 5173


31 7

Bunnett, J. F., Creary, X. and Sundberg, J. E. (1976). J. Org. Chem. 41, 1707 Cantacuzene, D. and Dorme, R. (1975). Tetrahedron Lett. 2037 Carruthers, R. A., Crellin, R. A. and Ledwith, A. (1969). J. Chem. SOC.,Chem. Commun. 252 Cedheim, L. and Eberson, D. (1976). Acta Chem. Scand. B30, 527 Chanon, M. (1982). Bull. SOC.Chim. Fr. 197 Chanon, M. (1985). Bull. SOC.Chim. Fr. 209 Chung, S.-K. (1980). J. Org. Chem. 45, 3513 Chung, S.-K. and Chung, F.-F. (1979). Tetrahedron Lett. 2473 Chung, S.-K. and Filmore, K. L. (1983). J. Chem. SOC.,Chem. Commun. 358 Citterio, A., Minisci, F., Porta, 0. and Sesana, G . (1977). J . Am. Chem. SOC.99, 7960 Clennan, E. L., Simmons, W. and Almgren, C. W. (1981). J. Am. Chem. SOC.103, 2098 Cohen, T., McMullen, C. H. and Smith, K. (1968). J. Am. Chem. SOC.90, 6866 Crozet, M. P. and Vanelle, P. (1985). Tetrahedron Lett. 323 Crozet, M. P., Surzur, J.-M., Vanelle, P., Ghiglione, C. and Maldonada, J. (1985). Tetrahedron Lett. 1023 Darchen, A., Mahe, C. and Patin, H. (1982). J. Chem. SOC.,Chem. Commun. 243 Dass, C. and Gross, M. L. (1984). J. Am. Chem. SOC.106, 5775 Delaunay, J., Lebouc, A., Tallec, A. and Simonet, J. (1982). J. Chem. SOC.,Chem. Commun. 387 Deno, N. C., Billups, W. E., Fishbein, R., Pierson, C., Whalen, R. and Wyckoff, J. C. (1971). J . Am. Chem. SOC.93,438 Detar, D. F. and Turetzky, M. N. (1955). J. Am. Chem. SOC.77, 1745 Detar, D. F. and Turetzky, M. N. (1956). J. Am. Chem. SOC.78, 3925, 3928 Dickerman, S. C. and Weiss, K. (1957). J. Org. Chem. 22, 1020 Dickerman, S. C., Weiss, K. and Ingberman, A. K. (1956). J. Org. Chem. 21, 380 Dickerman, S. C., Weiss, K. and Ingberman, A. K. (1958). J . Am. Chem. SOC.80, 1904 Eberson, L. (1982). Adv. Phys. Org. Chem. 18, 79 Eberson, L. (1983). J. Mol. Cat. 20, 27 Eberson, L. and Jonsson, L. (1980). J. Chem. SOC.,Chem. Commun. 1187 Farid, S. and Shealer, S. E. (1973). J. Chem. SOC.,Chem. Commun. 677 Feiring, A. E. (1983). J. Org. Chem. 48, 347 Feiring, A. E. (1985). J. Org. Chem. 50, 3269 Freeman, D. J. and Norris, R. K. (1976). Austral. J . Chem. 29, 2631 Freeman, D. J., Norris, R. K. and Woolfenden, S. K. (1978). Austral. J . Chem. 31, 2477 Gardini, G . P. and Bargon, J. (1980). J . Chem. SOC.,Chem. Commun. 757 Garst, J. F. and Barton, F. E. (1969). Tetrahedron Lett. 587 Giese, B. (1983). Angew. Chem. Int. Ed. EngI. 22, 753 Giese, B. (1985). Angew. Chem. Int. Ed. Engl. 24, 553 Giese, B. and Groningen, K. (1984). Tetrahedron Lett. 2473 Giese, B. and Horler, H. (1985). Tetrahedron 41, 4025 Gilbert, B. C. and Norman, R. 0. C. (1982). Can. J. Chem. 60, 1379 Gilbert, B. C., Holmes, R. G. G., Laue, H. A. H. and Norman, R. 0. C. (1976). J . Chem. Soc.. Perkin. Trans. 2 1047 Gilbert, B. C., Holmes, R. G. G. and Norman, R. 0. C. (1977). J . Chem. Res. ( M ) 1001 Girdler, D. J. and Norris, R. K. (1975). Tetrahedron Lett. 2375

31 8

G L E N A. RUSSELL

Groenewold, G. S. and Gross, M. L. (1984a). J. Am. Chem. SOC.106, 6575 Groenewold, G. S. and Gross, M. L. (1984b). J. Am. Chem. SOC.106, 6569 Groenewold, G. S., Chess, E. K. and Gross, M. L. (1984). J. Am. Chem. SOC.106, 537 Groves, J. T. and Ma, K. W. (1974). J. Am. Chem. SOC.96, 6527 Hatem, J. and Waegell, B. (1973). Tetrahedron Lett. 2023 Hayes, R. K., Roberts, M. K. S. and Wilmot, I. D. (1978). Austral. J. Chem. 31, 1737 Hegedus. L. S. and Miller, L. L. (1975). J. Am. Chem. SOC.97, 459 Hegedus, L. S. and Thompson, D. H. P. (1985) J. Am. Chem. SOC.107, 5663 Heiba, E. I., Dessau, R. M. and Koehl, W. J. (1969). J. Am. Chem. SOC.91, 6830 Hershberger, J. W. and Kochi, J. K. (1982). J. Chem. SOC.,Chem. Commun.212 Hixson. S. S. (1979). Org. Photochem. 4, 191 Hixson, S. S., Boyer, J. and Gallucci, C. (1974). J. Chem. SOC.,Chem. Commun.540 Hofmann, A. W. (1883). Chem. Ber. 16, 558, 586 Horner, L. and Anders, B. (1962). Chem. Ber. 95, 2470 Horner, L. and Betzel, C. (1953). Justus Liebigs Ann. Chem. 175, 579 Horner, L. and Schwenk, E. (1944). Angew Chem. 61, 411 Hrabak, F. and Lokaj, J. (1970). Collect. Czech. Chem. Commun. 35, 1081 Huyser, E. S. and Harmony, J. A. K. (1974). In “Methods in Free-Radical Chemistry” (ed. E. S. Huyser), Vol 5, p. 101. Marcel Dekker, Inc., New York Huyser, E. S. and Kahl, A. A. (1970). J. Org. Chem. 35, 3742 Huyser, E. S., Bredeweg, C. J. and VanScoy, R. M. (1964). J. Am. Chem. SOC.86, 4148 Huyser, E. S., Harmony, J. A. K. and McMillian, F. L. (1972). J . Am. Chem. SOC.94, 3176 Jawdosiuk, M., Makosza, M., Mudryk, B. and Russell, G. A. (1979). J . Chem. Soc., Chem. Commun. 488 Jugelt, W. and Pragst, F. (1968). Angew Chem. Int. Ed. 7, 290 Kharasch, M. S. and Sosnovsky, G. (1958). J. Am. Chem. SOC.80, 765 Kharasch, M. S., Pauson, P. and Nudenberg, W. (1953). J . Org. Chem. 18, 332 Kill, R. G. and Widdowson, D. A. (1976). J. Chem. SOC.,Chem. Commun. 755 Kim, J. K. and Bunnett, J. F. (1970). J . Am. Chem. SOC.92, 7463, 7464 Kinney, R. J., Jones, W. D. and Bergman, R. G. (1978). J. Am. Chem. SOC.100,635 Kochi, J. K. (1957). J. Am. Chem. SOC.79, 2942 Kochi, J. K. (1967). Science 155, 415 Koelsch, C. F. and Boekelheide, V. (1944). J. Am. Chem. SOC.66,412 Kornblum, N. (1975). Angew. Chem. [Int. Ed. Engl.] 14, 734 Kornblum, N. (1982). In “The Chemistry of Functional Groups. Supplement F” (ed. S. Patai), Ch. 10, p. 361. John Wiley and Sons, New York Kornblum, N. and Boyd, S. D. (1970). J. Am. Chem. SOC.92, 5784 Kornblum, N. and Cheng, L. (1977). J. Am. Chem. SOC.42, 2944 Kornblum, N. and Erickson, A. S. (1981). J . Org. Chem. 46, 1037 Kornblum, N. and Fifolt, M. J. (1980). J. Org. Chem. 45, 360 Kornblum, N. and Stuchal, F. W. (1970). J. Am. Chem. SOC.92, 1804 Kornblum, N., Michael, R. E. and Kerber, R. C. (1966). J. Am. Chem. SOC.80, 5660, 5662 Kornblum, N., Davies, T. M., Earl, G. W., Green, G. S., Holy, N. L., Kerber, R. C., Manthey, J. W., Musser, M. T. and Snow, D. H. (1967a). J. Am. Chem. SOC. 89, 725 Kornblum, N., Davies, T. M., Earl, G. W., Holy, N. L., Kerber, R. C., Musser, M. T. and Snow, D. H. (1967b). J. Am. Chem. SOC.89, 5714 Kornblum, N., Davies, T. M., Earl, G. W., Pinnick, H. W. and Stuchal, F. W. (1968). J. Am. Chem. SOC.90, 6219


31 9

Kornblum, N., Boyd, S. D. and Stuchal, F. W. (1970a). J. Am. Chem. Soc. 92, 5783 Kornblum, N., Swiger, R. T., Earl, G. W., Pinnick, H. W. and Stuchal, F. W. (1970b). J . Am. Chem. Soc. 92, 5513 Kornblum, N., Boyd, S. D., Pinnick, H. W. and Smith, R. G. (1971). J . Am. Chem. Soc. 93, 4316 Kornblum, N., Kestner, M. M. Boyd, S. C., and Cattran, L. C. (1973). J . Am. Chem. Soc. 95, 3356 Kornblum, N., Boyd, S. D. and Ono, N. (1974). J . Am. Chem. SOC.96, 2580 Kornblum, N., Carlson, S. C. and Smith, R. G. ( 1 978a). J. Am. Chem. Soc. 100,289 Kornblum, N., Carlson, S. C., Widner, J., Fifolt, M. J., Newton, B. N. and Smith, R. G. (1978b). J . Org. Chem. 43, 1394 Kornblum, N., Carlson, S. C. and Smith, R. G . (1979a). J . Am. Chem. Soc. 101, 647 Kornblum, N., Widmer, J. and Carlson, S. C. (1979b). J . Am. Chem. Soc. 101, 658 Kornblum, N., Ackerman, P. and Swiger, R. T. (1980). J . Org. Chem. 45, 5294 Kornblum, N., Singh, H. K. and Boyd, S. D. (1984). J . Org. Chem. 49, 358 Korzan, D. G., Chen, F. and Ainsworth, C. (1971). J . Chem. Soc. Chem. Commun. 1053 Kossi, R., Simonet, J. and Dauphin, G. (1980). Tetrahedron Lett. 3575 Kurosawa, H., Okada, H. and Yasuda, M. (1980). Tetrahedron Lett. 959 Kurosawa, H., Okada, H. and Hattori, T. (1981). Tetrahedron Lett. 4495 Lamb, R. C., Ayers, P. W., Toney, M. K. and Garst, J. F. (1966). J . Am. Chem. Soc. 88,4261 Ledwith, A. (1972). Acc. Chem. Res. 5, 133 Loffler, K. and Freytag, C. (1909). Chem. Ber. 42, 3427 Lu, J.-J. and Truce, W. E. (1985). J . Org. Chem. 50, 4885 Majima, T., Pac, C., Nakasone, A. and Sakurai, H. (1978). J. Chem. Soc., Chem. Commun. 490 Majima, T., Pac, C. and Sakurai, H. (1980). J . Am. Chem. Soc. 102, 5265 Malhotra, S. K., Hostynek, J. J. and Lundin, A. F. (1968). J . Am. Chem. SOC.90, 6565 Mattes, S. L. and Farid, S. (1982). Acc. Chem. Res. 15, 80 Mayake, H. and Yamamura, K. (1986). Bull. Chem. Soc. Jpn. 59, 89 Meijs, G. F. (1984). J . Org. Chem. 49, 3863 Meijs, G. F. (1985). Tetrahedron Lett. 105 Meijs, G. F. (1986). J. Org. Chem. 51, 606 Meyers, C. Y. (1978). “Topics in Organic Sulfur Chemistry”, p. 207. University Press, Ljubljana, Yugoslavia Meyers, C. Y., Matthews, W. S., Ho, L. L., Kolb, V. M. and Parady, T. E. (1977). In “Catalysis in Organic Synthesis”, p. 197. Academic Press, New York Minisci, F., Galli, R. and Cecere, M. (1965). Tetrahedron Letf. 4663 Minisci, F., Citterio, A,, Vismara, E. and Giordano, C. (1985). Tetrahedron 41,4170 Nelsen, S . F. and Teasley, M. F. (1986). J. Org. Chem., 51, 3221 Nelsen, S. F., Kapp, D. L. and Teasley, M. F. (1984). J . Org. Chem. 49, 579 Nelsen, S. F., Kapp, D. L., Gerson, F. and Lopez, J. (1986). J . Am. Chem. SOC.108, 1027 Neunteufel, R. A. and Arnold, D. R. (1973). J . Am. Chem. So;. 95, 4080 Newcombe, P. J. and Norris, R. K. (1978). Austral. J. Chem. 31, 2463 Nonhebel, D. C. and Waters, W. A. (1957). Proc. Roy. Soc., A 242, 16 Norris, R. K. (1983). In “The Chemistry of Functional Groups, Supplement D”, (eds S. Patai and Z. Rappoport), Ch. 16, p. 681. John Wiley and Sons, New York Norris, R. K. and Randles, D. (1976). Austral. J . Chem. 29, 2621 Norris, R. K. and Randles, D. (1982a). Austral. J . Chem. 35, 1621 Norris, R. K. and Randles, D. (1982b). J . Org. Chem. 47, 1047

320

GLEN A. RUSSELL

Nugent, W. A. and Kochi, J. K. (1977). J. Organometal. Chem. 124, 327, 349, 371 Ohno, A., Shio, T., Yamamoto, H. and Oka, S. (1981). J. Am. Chem. SOC.103, 2045 Ono, N., Tamura, R. and Kaji, A. (1980). J. Am. Chem. SOC.102, 2851 Ono, N., Tamura, R., Eto, H., Hamamoto, I., Nakatuska, T., Hayami, J.-I. and Kaji, A. (1983). J. Org. Chem. 48, 3678 Ono, N., Miyake, H., Kamimura, A., Hamamoto, I., Tamura, R. and Kaji, A. (1985). Tetrahedron 41, 4013 Ostovic, D., Roberts, R. M. G. and Kreevoy, M. M. (1983). J . Am. Chem. Soc. 105, 7629 Pabon, R. A. and Bauld, N. L. (1983). J. Am. Chem. SOC.105, 633 Pabon, R. A. and Bauld, N. L. (1984). J. Am. Chem. SOC.106, 1145 Pabon, R. A., Bellville, D. J. and Bauld, N. L. (1983). J. Am. Chem. SOC.105, 5158 Pabon, R. A., Bellville, D. J. and Bauld, N. L. (1984). J. Am. Chem. SOC.106, 2730 Pac, C., Nakasone, A. and Sakurai, H. (1977). J. Am. Chem. SOC.99, 5866 Palacios, S. M., Santiago, A. N. and Rossi, R. A. (1984). J . Org. Chem. 49, 4609 Palacios, S. M., Alonso, R. A. and Rossi, R. A. (1985). Tetrahedron 41, 4147 Pierini, A. B., Penenory, A. B. and Rossi, R. A. (1985). J. Org. Chem. 50, 2739 Popov, V. I., Boiko, V. N. and Yagupol'skii, L. M. (1977). J. Org. Chem. USSR, [Engl. Trans.] 13, 1985 Pross, A. (1985). Acc. Cherti. Res. 19, 212 Prousek, J. (1980). Czech. Chem. Commun. 45, 3347 Prousek, J. (1982). Czech. Chem. Commun. 47, 1334 Razuvaev, R. A,, Petukhov, G . G., Zhil'tsov, S. F. and Kudryavtsev, L. F. (1960). Dokl, Akad. Sci. 135, 87 Rico, I., Cantacuzkne, D. and Wakselman, C. (1981). Tetrahedron Lett. 3405 Rico, I., Cantacuzhe, D. and Wakselman, C. (1983). J . Org. Chem. 48, 1979 Rossi, R. A,, Palacios, S. M. and Santiago, A. N. (1982). J. Org. Chem. 47, 4654 Rossi, R. A,, Santiago, A. N. and Palacios, S. M. (1984). J . Org. Chem. 49, 3387 Roth, H. D. and Schilling, M. L. M. (1979). J. Am. Chem. SOC.101, 1898 Roth, H. D. and Schilling, M. L. M. (1980a). J . Am. Chem. SOC.102, 4303 Roth, H. D. and Schilling, M. L. M. (1980b). J. Am. Chem. SOC.102, 7956 Roth, H. D. and Schilling, M. L. M. (1981). J. Am. Chem. SOC.103, 7210 Russell, G . A. (1953). J. Am. Chem. SOC.75, 5100 Russell, G . A. (1959). J. Am. Chem. SOC.81, 4815, 4825 Russell, G. A. (1967). Pure and Appl. Chem. 15, 185 Russell, G. A. ( 1 982). Chemia Stosowana 26, 3 17 Russell, G. A. (1986). Preprints of Papers, Div. of Petroleum Chem., p. 891. 192nd National Meeting of the Am. Chem. SOC.,Anaheim, California Russell, G . A. and Bemis, A. G. (1965). Chem. and Znd. (London) 1262 Russell, G. A. and Danen, W. C. (1966). J. Am. Chem. SOC.88, 5663 Russell, G. A. and Danen, W. C. (1968). J. Am. Chem. SOC.90, 347 Russell, G . A. and Dedolph, D. F. (1985). J . Org. Chem. 50, 2378 Russell, G . A. and Guo, D. (1984). Tetrahedron Lett 5239 Russell, G . A. and Hershberger, J. (1980a). J. Chem. SOC.,Chem Commun. 216 Russell, G . A. and Hershberger, J. (1980b). J. Am. Chem. SOC.102, 7603 Russell, G. A. and Kaupp, G . (1969). J . Am. Chem. SOC.91, 3851 Russell, G. A. and Khanna, R. K. (1985a). J. Am. Chem. SOC.107, 1450 Russell, G. A. and Khanna, R. K. (1985b). Tetrahedron 41, 4133 Russcll, G. A. and Khanna, R. K. (1987). Adv. in Chem. Ser., Am. Chem. SOC.215, in press


321

Russell, G. A. and Ngoviwatchai, P. (1985). Tetrahedron Lett. 4975 Russell, G. A. and Pecoraro, J. M. (1979). J. Am. Chem. SOC.101, 3331 Russell, G. A. and Ros, F. (1985). J. Am. Chem. SOC.107, 2506 Russell, G. A. and Tashtoush, H. (1983). J. Am. Chem. SOC.105, 1398 Russell, G. A. and Weiner, S. A. (1969). Chem. and Ind. (London) 659 Russell, G. A., Janzen, E. G., Becker, H.-D. and Smentowski. F. (1962a). J . Am. Chem. SOC.84, 2652 Russell, G. A., Moye, A. J , and Nagpal, K. (1962b). J . Am. Chem. SOC.84,4154 Russell, G. A,, Janzen, E. G . and Strom, E. T. (1964). J . Am. Chem. SOC.86, 1807 Russell, G. A,, Geels, E. J., Moye, A. J., Janzen, E. G., Bemis, A. G., Mak, S. and Strom, E. T. (1965). Adv. in Chem. Ser., Am. Chem. SOC.51, 112 Russell, G. A., Bemis, A. G,,Geels, E. J., Janzen, E. G., and Moye, A. J. (1968). Adv. in Chem. Ser., Am. Chern. SOC.75, 174 Russell, G. A,, Norris, R. K. and Panek, E. J. (1971). J. Am. Chem. SOC.93, 5839 Russell, G. A., Hershberger, J. and Owens, K. (1979a). J. Am. Chem. SOC.101, 1312 Russell, G. A., Jawdosiuk, M. and Makosza, M. (1979b). J. Am. Chem. SOC.101, 2355 Russell, G. A., Jawdosiuk, M. and Ros, F. (1979~).J. Am. Chem. SOC.101, 3378 Russell, G. A., Ros, F. and Mudryk, B. (1980). J . Am. Chem. SOC.101, 7601 Russell, G. A,, Mudryk, B. and Jawdosiuk, M. (1981a). J . Am. Chem. SOC.103,4610 Russell, G. A., Mudryk, B. and Jawdosiuk, M. (1981b). Synthesis 62 Russell, G. A., Hershberger, J. and Owens, K. (1982a). J . Organomet. Chem. 225,43 Russell, G. A., Mudryk, B.. Jawdosiuk, M. and Wrobel, Z. (1982b). J. Org. Chem. 47, 1879 Russell, G. A., Mudryk, B., Ros, F. and Jawdosiuk, M. (1982~).Tetrahedron 38, 1059 Russell, G. A., Ros, F., Hershberger, J. and Tashtoush, H. (1982d). J . Org. Chem. 47, 4316 Russell, G. A., Tashtoush, H. and Ngoviwatchai, P. (1984). J. Am. Chem. SOC.106, 4622 Russell, G. A., Khanna, R. K. and Guo, D. (1985). J. Org. Chem. 50, 3423 Russell, G. A., Jiang, W. and Hu, S. S. (1986a). J. Org. Chem. 51, 5499 Russell, G. A., Khanna, R. K. and Guo, D. (1986b). J. Chem. SOC.,Chem. Commun. 632 Simig, G. and Lempert, K. (1961). Chem. Ber. 61, 607 Simig, G., Lempert, K., Tamas, J. and Szepsey, P. (1977). Tetrahedron Lett. 1151 S i m k G., LemPert, K., Vali, z.. Toth, G . and Tamas, J. (1978). Tefrahedron 34, 2371 Singh, P. R. and Jayaraman, B. (1974). Indian J . Chem. 12, 1306 Singh, P. R. and Khanna, R. K. (1983). Tetrahedron Lett. 141 1 Sin&, p. R., Nigham, A. and Khurana, J. M. (1980). Tetrahedron Lett. 4753 Singh, P. R., Khurana, J. M. and Nigham, A. (1981). Tetrahedron Lett. 2901 Spanswick, J. and Ingold, I(.U. (1970). Can. J. Chem. 48, 546 Tang, R., Yue, H. J., Wolf, J. F. and Mares, F. (1978). J. Am. Chem. SOC.100, 5248 Tanner, D. D. and Mosher, M. W. (1978). Can J. Chem. 47, 715 Tanner, D. D., Blackburn, E. V. and Diaz, G. E. (1981). J. Am. Chem. SOC.103, 1557 Tanner, D. D., Arhart, R. and Meintzer, C. (1985a). Tetrahedron 41, 4261 Tanner, D. D., Diaz, G. E. and Potter, A. (1985b). J. Org. Chem. 50, 2149 Tanner, D. D.3 Singh, H. K. and Yang, D. (1986a) Rev. Chem. Inrermed. 7, 13

322

GLEN A. RUSSELL

Tanner, D. D., Singh, H. K., Yang, D. and Stein, A. R. (1986b). Preprints of Papers, Div. qf’PefroleumChem., p. 897. 192nd National Meeting of the Am. Chem. SOC., Anaheim, California Walling, C. and Buckler, S. A. (1955). J. Am. Chem. SOC.77,6032 Walling, C. and Indictor, N. (1958). J . Am. Chem. SOC.80, 5814 Waters, W. A. (1942). J. Chem. SOC.266 Wawzonek, S. and Thelan, P. J. (1950). J . Am. Chem. SOC.72, 21 18 Wayner, D. D. M. and Arnold, D. R. (1985). Can. J . Chem. 63, 871 Wong, P. C. and Arnold, D. R. (1979). Tetrahedron Lett. 2101 Yasufuka, K. and Kutal, C. (1984). Tetrahedron Lett. 4893 Zeilstra, J. J. and Engberts, J. B. F. N. (1973). Red. Trav. Chim. Pays-Bas 92,954 Zeldrin, L. and Shechter, H. (1957). J . Am. Chem. SOC.79,4708 Zorin, V. V., Kukovitskii, D. M., Zlotskii, S. S., Todres, Z. V. and Rakhomankulov, D. L. (1983). J . Org. Chem. SSSR 53, 906

Author Index Numbers in italic refer to the pages on which references are listed at the end of each article

Abraham, E. P., 166, 202, 250, 261, 265, 267 Abrahamsson, S., 187, 188, 261 Ackerman, P., 280, 319 Adams, H., 45,58 Adriaens, P., 184, 261, 264 Agathocleous, D., 203, 205, 240, 250, 261 Ahlberg, P., 64, 93, 94, 95, 119, 158 Ainsworth, C., 297,319 Alder, H., 21, 58 Alder, R. W., 310,316 Aldwin, L., 254, 265 Aleixo, M. V., 225, 226, 263 Aleksandrov, Y. A., 275, 316 Al-Khalil, S. I., 277, 279, 316 Allinger, M. L., 197, 261 Almgren, C. W., 314,317 Alnaijar, M. S., 27, 61 Alonso, R. A., 286,320 Altman, L. J., 71, 82, 129, 159 Anders, B., 300, 318 Anderson, B. F., 188, 269 Anderson, B. M., 218, 262 Anderson, E. G., 252, 262 Anderson, R. L., 45, 59 Anderson, S. N., 19, 58 Ando, W., 314, 316 Andrews, D. W., 193,264 Andrews, G. C., 104, 162 Andrews, S. L., 190, 206, 267 Andrist, A. H., 65, 159 Anet, F. A. L., 95, 98, 102, 104, 106, 108, 158

Angelo, H. R., 201, 263 Anoardi, L., 231, 262 Applegate, H. E., 207, 261 Araujo, P. S., 225, 226, 263 Argyropolous, J. N., 286, 316 Arhart, R., 308,321

Arnold, D. R., 311, 314, 316, 319, 322 Aroney, M. J., 194, 262 Arrowsmith, C. H., 72, 160 Arvanaghi, M., 133, 141, 160 Ashby, E. C., 27, 58, 284, 286, 297, 298, 299, 316 Askani, R., 70, 93, 96, 137, 158 Asso, M., 221, 262 Aydin, R., 71, 72, 105, 158, 159 Ayers, P. W., 275, 319 Azam, K. A., 39, 58 Baborack, J. C., 92, 128, 161 Bagchler, R. D., 192, 269 Bailey. N. A., 45, 58 Baird, M. C . , 19, 62 Baker, W., 166, 261 Balch, A. L., 39, 58 Baldry, K. W., 74, 100, 101, 160, 161 Ballard, D. H., 19, 58 Baltzer, B., 259, 262 Baltzer, L. 72, 160 Banks, S., 284, 316 Barbetta, A., 16, 60 Bargon, J., 310, 317 Barker, S. D., 281, 284, 316 Barltrop, J. A., 299, 316 Bartolone, J. B. 180, 266 Barton, D. H. R., 313, 316 Barton, F. E., 297, 317 Bary, Y., 64, 159 Basus, V. J., 98, 102, 161 Batchelor, F. R., 233, 262 Batiz-Hernandez, H., 71, 158 Baudry, D., 50, 58 Bauld, N. L., 31 I , 316, 320 Beauchamp, J. L., 51, 60 Beaufays, F., 210, 267 Becker, H.-D., 275, 321 323

324

Becker, Y., 26, 58 Beeumen, J., van, 252, 266 Belasco, J. G . . 252, 264 Bellachioma, G . . 48, 58 Belluco, U.. 49, 62 Bellville. D. J., 31 I , 316, 320 Bemis, A. G., 274, 275, 320, 321 Benacia, K. E., 10, 61 Bender, M. L., 195, 232, 262, 266 Bennett, M. A., 36, 58 Berestova, S. S., 82, 161 Berezin, I. V., 223, 225, 226, 267 Berge, S. M., 207, 215, 262 Berger, S., 64, 72, 158, 160 Berger, S. A,, 170, 262 Bergman, R. G., 17, 30, 50, 53, 60, 61, 62, 297, 298, 318 Bernheim, R. A., 71, 158 Bethell, D., 309, 316 Betzel, C., 300, 318 Beugelmans, R.,278, 281, 316 Biali. S. E., 75, 160 Bigeleisen, J., 66, 158 Billups, W. E., 308, 317 Binkley, J. S., 186, 266 Bird, A. E., 258, 261, 262 Birk, J. P., 33, 58 Birladeanu, L., 193, 264 Bisnette, M. B., 44,60 Blackburn, E. V., 304, 316, 321 Blackburn, G. M., 196, 197, 233, 262 Blaha, J. M., 209, 212, 262 Blake, D. M., 36, 49, 61, 62 Block, P. L., 20, 21, 27, 58, 59 Blum, J., 33, 42, 59, 61 Blumel, P., 174, 268 Boeck, L. D., 166, 265, 267 Boekelheide, V., 272, 318 Bogachev, Y. S., 82, 161 Bogden, S., 240, 268 Boiko, V. N., 285, 316, 320 Bolton, P. D., 21 I , 262 Booth, B. L., 12, 59 Booth, H., 74, 75, 101, 158 BorEiC, S., 87, 137, 162 Bordwell, F. G., 280, 295,316 Borgese, J., 202, 270 Borsub, N., 312,316 Boschetto, D. J., 19, 20, 59, 62 Bose, A. K., 189, 197, 262, 267

AUTHOR INDEX

Bosnich, B., 40, 62 Botkin, J. H., 72, 159 Bottini, A. T., 194, 262 Bouchoux, G., 210, 262 Bowers, M . T., 51, 60 Bowman, N. S.. 65, 69. 159 Bowman, W. R., 277,278,279,316 Boyd, D. B., 186, 191, 192, 202, 206, 250, 262, 268 Boyd, S. D., 277, 279, 282, 296, 318, 319 Boyer, J., 314, 318 Boyle, W. J., 298, 299, 316 Bradbury, D., 299,316 Bradley, J. s.,23, 59 Brant, S. R.,254, 266 Brauman, J. I., 14, 17, 38, 42, 59 Braun, S., 64, 160 Braus, R. J, 23,61 Bredeweg, C. J., 300,318 Brittain, W. J., 142, 144, 146, 160 Brookhart, M., 117, 158 Brown, H. C., 123, 135, 158, 197, 262 Brown, M. P., 39, 58 Brown, R. K., 118, 162 Brown, T. L., 16, 61 Brownstein, S., 88, 158 Broxton, T. J., 231, 262 Bruce, M. I., 294,316 Bruckner, D., 89, 158 Brunet, J. J., 286, 316 Bruylants, A., 199, 21 I , 262 Buchanan, D. H., 19, 21, 60 Buchanan, M., 53, 62 Buckler, S. A., 275, 322 Buckwalter, F. H., 215, 269 Buckwell, S., 203, 205, 240, 250, 251, 261, 262 Bucourt, R.,207, 262 Bulm, G., 233, 264 Bundgaard, H., 198, 201, 202, 203, 207, 209, 212, 214, 215, 218, 233, 234, 236, 240, 248, 249, 250, 253, 254, 262, 263, 266 Bunn, C. W., 166,263 Bunnett, J. F., 276, 285, 298, 299, 316, 317, 318 Bunton, C. A., 223, 225, 263 Burgen, A. S. V., 182, 209, 213, 263 Burns, D., 16, 59

AUTHOR INDEX

Busch, D. H., 41, 61 Buss, V., 193, 263 Busson, R., 191, 258, 263 Butler, A. R., 248, 263 Buur, A., 234, 263 Bywater, S., 88, 158 Calabrese, J. C., 38, 61 Calderazzo, F., 19, 59 Calvert, R. B., 109, 161, 162 Cantacuzene, D., 285, 301, 317, 320 Cardaci, G., 48, 58 Carlsen, N. R., 192, 263 Carlson, S. C., 279, 280, 283, 297.,298, 319 Carroll, R. D., 258, 263 Carruthers, R. A., 310, 317 Cartwright, S. J., 252, 263 Casey, C. P., 45, 59, 1 12, 158 Casson, A., 310, 316 Cattran, L. C., 277, 319 Caubere, P., 286,316 Caulton, K . G., 50, 62 Cawse, J . N., 14, 17, 38, 42, 59 Cecere, M., 308, 319 Cedheim, L., 311,317 Cenini, S., 11, 15, 16, 44, 48, 62 Chaiet, L., 166, 267 Chaimovich, H., 225, 226, 263 Chain, E., 166, 261 Chakrawarti, P. B., 221, 263 Chambers, R., 197, 263 Chan, S. I., 82, 158 Chandrasekhar, J., 64, 75, 161 Chanjamsri, S., 16, 60 Chanon, M., 315,317 Chari, S., 92, 128, 161 Charlier, P., 177, 265 Charsley, C.-H., 261, 262 Chatt, J., 51, 59 Chelsky, R., 311, 316 Chen, F., 297, 319 Cheney, A. J., 36,59 Cheng, A. K., 95, 104, 158 Cheng, L., 278, 296, 318 Chess, E. K., 312, 318 Chiang, Y., 210, 266 Chipperfield, J. R.,7, 60 Chmurny, G. N . , 104, 162

325

Chock, P. B., 10, 14, 15, 59 Christensen, B. G., 206, 259, 261, 264 Christ], M., 89, 158 Chrzastowski, J. Z., 19, 58 Chung, F.-F., 297, 299,317 Chung, S.-K., 297, 298, 299, 317 Cimarusti, C. M., 182, 240, 265, 270 Citterio, A,, 273, 302, 317, 319 Claes, P. J., 182, 191, 263, 265 Clark, D. R., 54, 59 Clayton, J. P., 261, 263 Clemens, A. H., 280, 295,316 Clennan, E. L., 314, 317 Clutter, D., 82, 158 Coates, R. M., 133, 159 Coene, B., 189, 203, 206, 263, 264,268, 269 Cohen, S. A., 252, 263 Cohen, T., 272, 317 Cole, T. E., 54, 59 Coleman, A. W., 41, 59 Collings, A. J., 191, 263 Collins, C. J . , 65, 69, 159 Collman, J. P., 7, 10, 14, 17, 36, 37, 38, 42, 54, 59 Coluisio, J. T., 218, 220, 263 Colville, N. J., 24, 61 Connor, D. E., 23,59 Considine, J. L., 28, 61 Cooke, M. P., 54, 59 Cooksey, C. J., 30, 59 Copperthwaite, R. G., 3, 61 Coppola, G. M., 84, 161 Cordes, E. H., 218, 223, 262, 263 Correia, V . R., 225, 226, 263 Cotton, F. A., 5, 59, 73, I59 Coyette, J., 177, 265 Crabtree, R. H., 3, 50, 53, 59 Cram, D. J., 193, 264 Creary, X., 285, 31 7 Cree-Uchiyama, M., 11 I , 162 Crellin, R. A., 310, 317 Cressman, W. A., 218, 220, 263 Crombie, D. A., 232, 268 Crowfoot, D., 166, 263 Crozet, M. P., 278, 280, 281, 317 Cryberg, R. L., 194, 265 Cuccovia, I. M., 225, 226, 263 Cushman, M., 209, 210, 213, 258, 266 Cutmore, E. A., 258,262

326

AUTHOR INDEX

de Waal, D. J. A., 3, 38, 60, 61 Dewar, M. J. S., 129, 145, 159 Dewdney, J. M., 233, 262 DeWeck, A. L., 233,253, 264,269 Dahl, L. F., 186, 187, 188, 190, 201, Dewhirst, K. C., 193, 264 202, 206,269 DeWit, D. G., 50, 62 Damon, R., 84,161 Dhami, K. S., 189, 264 Danen, W. C., 276, 277, 280,320 Diaz, G. E., 304, 305, 321 D'Antonio, P., 82, 160 Dickerman, S. C., 272,317 Darchen, A., 294, 317 DiCosimo, R., 52,60 Darensbourg, D. J., 16, 59 Dideberg, O., 177, 265 Darensbourg, M. Y.,16, 17, 59 Diehl, B. W. K., 72, 158 Dass, C., 312, 317 Dinner, A., 249 264 Dauphin, G., 309, 319 Davidson, J. M., 51, 59 Dive, G., 177, 265 Davies, T. M., 280, 318 Dodd, D., 19, 30, 58, '59 Davis, A. M., 252, 254, 255, 256, 258, Dodds, H. L. H., 197,262 Dodrell, D., 27, 28, 61 259, 263 Doedens, R. J., 197,263 Davis, D. D., 19, 27, 60 Doering, W. von E., 193, 264 Davis, N. E., 166, 265 Dolfini, J. E., 240, 270 Davis, W. W., 209, 264 Dolphin, D., 23, 59, 61 Dawkins, G. M., 112, 162 Domenick, R. L., 72, 161 Dea, P., 82, 158 Donovan. D. J... 141.. 146.. 160 Deavers, J. P., 20, 59 Dorman, 'D. E., 189, 190, 206, 268 Dedobh. D. F., 280, 282, 320 Deeming, A. J.,'4, lo, 34, 35, 36, 59, 60 Dorme, R., 301, 317 DeFrees, D. J., 146, 157, 159, 186, 266 Dovek, I. C., 14, 42, 60 Douglas, A. W., 250, 265 Deganello, G., 49, 62 Doyle, F. P., 198, 264 Degelaen, J., 182, 184, 209, 213, 261, Drew, D. A., 16, 59 263, 264 Druzhkov, 0. N., 275, 316 DeGraeve, J., 184, 264 Ducep, J.-B., 191, 270 Deguchi, Y.,249, 269 Duddy, N. W., 231, 262 Dekmezian, A. H., 102, 104, 158 Duez, C., 177, 182, 265 Delaunay, J., 309, 317 Duncombe, R. E., 202,268 Delduce, A. J., 253, 269 Durant, F., 182, 207, 266 DeLigny, D. L., 232, 267 Dusart, J., 177, 180, 252, 264, 265, 266 Demanet, C. M., 3,61 Duus, F., 83, 84, 159 Demarco, P. V., 190, 263 Dennen, D. W., 209, 264 Deno, N. C., 308,317 Eadie, D. T., 41, 59 DePriest, R. N., 27, 58, 297, 298, 299, Earl, G. W., 280, 318,319 316 Earl, H. A., 195, 264 DePuy, C. H., 65,159 Eberson. D.. 31 1. 317 De Rango, C., 197,266 Dereppe, J. M., 189, 203, 206, 263, 264, Eberson, L., 308, 315, 317 Edgell, W. F., 16, 60 268, 269 Deslongchamps, P., 243, 264 Eichelberger, H. R., 220, 264 Eichorn, G., 220, 264 Dessau, R. M., 308, 318 Dessy, R. E., 12, 14, 60 Eigen, M., 238, 264 Eisen, H. N., 233, 268 Detar, D. F., 299, 317 Deutsch, E., 8, 12, 14, 41,42, 62 Emanuel, E. L., 252, 266

Cyr, C. R., 45,59

327

AUTHOR INDEX

Engberts, J . B. F. N., 277, 322 Engdahl, C., 64, 93, 94, 119, 158 Ephritikhine, M., 50, 58 Erickson, A. S., 279, 282, 318 Ernst, L., 72, I59 Espenson, J . H., 42, 62 Eto, H., 277, 320 Euler, K., 97, 160 Everett, J . R., 74, 75, 101, 158 Evilia. R. F., 64, 161 Eyssen, H., 184, 261 Ezumi, K., 206, 207, 267 Fagan, P. J., 28, 62, 112, 158 Fahey, J. L., 206, 264 Fairhurst, S. A., 309, 316 Fallab, S., 220, 270 Faller, J . W., 85, 161 Fanelli, V., 189. 267 Faraci, W. S., 250, 264 Farid, S., 311, 317, 319 Farr, J . P., 39, 58 Fauvarque, J . F., 33, 43, 49, 60 Fazakerley, G. V., 220, 221, 264 Feeney, J.. 182, 209, 213, 263 Feher, F. J., 52, 60 Feiring, A. E., 279, 283, 285, 31 7 Felkin, H., 50, 58 Fendlcr, E. J., 223, 264 Fendler, J. H., 223, 264 Fendrich, G., 254, 266 Fernandez, G. M., 215, 216, 266 Fersht, A. R., 184, 264 Fifolt, M . J., 279, 285, 318, 319 Figdore, P. E., 14, 16, 48, 61 Filmore, K. L., 299, 317 Finholt, P., 214, 264 Fink, A. L., 252, 263 Finke, R. G., 14, 17, 38, 41, 42, 59, 60 Firestone, R. A,, 206, 259, 261, 264 Fishbein, R., 308, 317 Fisher, J., 252, 264 Fitton, P., 42, 43, 60 Fitzgerald, P. H., 210, 266 Fleming, A,, 166, 264 Flynn, E. H., 189, 205, 264 Fong, F. K., 129, 159 Forsen, S., 71, 82, 129, 159 Forster, D., 38, 39, 56, 60

Forsyth, D. A,, 64. 72. 148, 159 Frank, M . J., 207, 215. 262 Franks, S., 7, 60 Freeman. D. J . , 284, 285, 31 7 Frere, J . - M . , 177, 180, 182, 184, 207, 252, 262, 263, 264, 265, 266 Fretz, E. R., 133, 159 Freytag, C., 308. 319 Fries, R. W., 25, 61 Fujimoto. M . , 33, 61 Fujita, T., 215, 265 Funderburk. L., 254. 265 Fiinfschilling, P. C., 65, 159 Fusi. A., 1 I , 15, 16, 44, 48. 62 Gagnon, J., 252, 266 Gajda, G. J., 91, 159 Galli, R., 308, 319 Gallucci, C., 3 14, 318 Gandemer, A., 31, 62 Gandler, J . R., 254, 266 Gani, V., 231, 265 Gardini, G . P., 310, 317 Gardner. S., 311, 316 Garibay, M. E., 72, 159 Garlenmeyer, H., 220, 270 Garst, J . F., 275, 297, 317, 319 Gartzke, W., 38, 60 Gassman, P. G., 194, 265 Gatford, C., 30, 59 Gaughan, G., 41,60 Gaylor, J. R., 10, 60 Gazzard, D., 233, 262 Geels, E. J., 274, 275, 321 Gensmantel, N. P., 193, 195, 196, 197, 199, 201, 202, 203, 206, 207, 209, 210, 211, 212, 213, 214, 215, 218, 219, 220, 221, 222, 223, 224, 225, 227, 228, 229, 230, 232, 233, 238, 239, 242, 243, 244, 245, 246, 247, 248, 254, 255, 258. 261,265, 268 George, C., 82, 160 George, G. M . St., 1 1 1 , 162 Georgopapadakou. N. H.. 182,265 Gerber, T. I. A,, 3, 38, 60, 61 Gerson, F., 314, 319 Ghebre-Sellassie, I., 209, 210, 213, 258, 266 Ghiglione, C., 281, 317

328

Ghosez, L., 240. 268 Ghuysen, J.-M., 177, 180, 182, 184, 207. 261,263, 264, 265,266 Giesbrecht, P., 174, 268 Giese, B., 290, 304, 317 Giffney, C. J., 212,265 Gilbert, B. C., 294, 317 Gilboa, H., 64, 159 Gilpin. M. L., 240, 265 Giordano, C., 302,319 Girdler, D. J., 284, 317 Gleiter, R., 193, 263 Glidewell, C., 191, 192, 265 Godfrey, J. C., 187, 201, 270 Goel, A. B., 284, 286, 297, 298, 299, 316 Gold, V., 64, 159, 254, 265 Goldberg, M., 189, 267 Good, M. L., 220, 264 Gorman, M., 166, 267 Goscinski, O., 95, 158 Gosteli, J., 190, 191, 201, 268 Gougoutas, J. Z . . 207, 261 Cowling, E. W., 193, 199, 201, 207, 215, 218, 222, 233, 244, 246, 247, 248, 265 Grabowski. E. J. J., 250. 265 Graham, W. A., 7, 9, 12, 14, 15, 32, 41, 42, 60, 61 Granatek. A. P., 215, 269 Gravitz, N., 254, 265 Green, G. G. F. H., 187, 265 Green, G. J., 31 I , 316 Green, G . S., 280, 318 Green, M., 112, 162 Green, M. L. H., 50, 60 Gregory, C. D., 14, 61 Gresser, M., 255, 267 Grimme, W., 93, 159 Grist, S., 254, 265 Grob, C. A,, 123, 159 Groenewold, G. S., 311, 312, 318 Groningen, K., 290, 317 Gross, M. L.. 311, 312, 317, 318 Groves, J. T., 297, 298, 318 Grubbs, R. H., 91, 159 Grundstrom, T., 252, 266 Guilian, M., 221, 262 Gunnarsson, G., 71, 82, 129, 159 Giinther, H., 71, 72, 93, 105, 158, 159

AUTHOR INDEX

Guo, D., 290, 291, 303, 307,320 Hagihara, T., 26, 60 Haiching, Z., 180, 266 Haines, L. M., 37, 60 Halevi, E. A., 64, 66, 72, 159 Hall, D., 187, 188, 266 Hall, M. L., 27, 28, 61 Halle. L. F., 51, 60 Halpern, J., 4, 10, 14, 15, 33, 48, 50, 58, 59, 60, 62 Hamamoto, I., 277, 300, 304,320 Harnano, S., 249,269 Harnill, R. L., 166, 265, 267 Hamilton-Miller, J. M. T., 202, 250, 265 Hammond, B. L., 125, 148, 160 Harnrnond, G. S., 193, 265 Hanckel, J. M., 16, 17, 59 Handler, A., 82, 161 Handoo, K . L., 309, 316 Hanratty, M. A., 51, 60 Hansen, H. J., 65, 159 Hansen, J., 218, 253, 263 Hansen, P. E., 64, 71, 83, 84, 159 Harbridge, J. B., 240, 265 Harlow,-R. L., 118, 162 Harmony, J. A. K., 300, 318 Harrod, J. F., 48, 60 Hart-Davis, A. J., 7, 9, 12, 14, 15, 41, 42, 60 Hartley, F. R., 7, 60 Hartshorn, S. R., 68, 70, 92, 137, 162 Haseltine, R., 125, 162 Hashimoto, N., 261, 265 Haszeldine, R. N., 12, 59 Hatern, J., 297, 318 Hattori, T., 300, 319 Hayami, J.-I., 277, 320 Hayashi, T., 26, 60 Hayes, R. K., 313, 316, 318 Hedge, S., 16, 60 Hegedus, L. S., 286, 318 Hehre, W. J., 68, 146, 152, 155, 157, 158, 159, 162, 193, 265 Heiba, E. I., 308, 318 Hem, S. L., 209, 210, 213, 258, 266 Henderson, N. L., 207, 215, 262 Hermann, R. B., 202, 262

329

AUTHOR INDEX

Hern, S. L., 209. 212, 262 Herron, D. K., 202. 206, 262 Hershberger, J., 277, 285, 293, 320, 321 Hershberger, J. W., 294, 318 Hewett, A. P. W., 98, 102, 161 Heymes, R., 207, 262 Hickey, C. M., 39, 48. 60 Hidalgo, A,, 221, 270 Higgins, C. E., 166, 265, 267 Hill, R. H., 4, 39, 58, 60 Hine, J., 254, 265 Hite, G. J., 180, 266 Hixson, S. S., 314, 318 Hlatky, G. G., 3, 59 Ho, L. L., 285, 319 Ho, P. P. K., 206, 265 Hodge, C. N., 120, 160 Hodgkin, D. C., 187, 188, 261, 266, 269 Hoehn, M. M., 166, 265, 267 Hoffmann, R., 3, 53, 62 Hofmann, A. W., 308, 318 Hogeveen, H., 134, 159 Holdrege, C. T., 187. 201, 270 Holmes. R. G. G., 294, 317 Holt, E. M., 53, 59 Holy, N. L., 280, 318 Hong, J. T., 224, 265 Hopf, H., 72, 159 Horiuchi, S., 249, 269 Horler, H., 290, 304, 317 Horner, L., 300, 318 Hostynek, J. J., 313, 319 Houriet, R., 51, 60, 210, 262 Hout, Jr. R. F., 68, 159 Howard, T. R., 91, 159 Howarth, 0. W., 116, 160 Howarth, T. F., 240, 265 Hrabak, F., 300, 318 Hu, S. S., 306, 321 Huang, E., 130, 162 Huang, M. B., 95, 158 Hull, W. W., 75, 160 Humphrey, M. B., 117, 158 Humphrey, Jr., J. S., 152, 162 Humphreys, D. W. R., 314, 316 Humski, K., 87, 162 Hunt, C. T., 39, 58 Hupe, D. J., 254, 266 Hursthouse, M. B., 47, 61 Huttner, G., 38, 60

Huyser. E. S., 300, 318 Ichikawa, K., 197, 262 Ikariya, T., 91, 159 Illies, A. J., 5 I , 60 Ilver, K., 215, 263 Indelicato, J. M., 190, 203, 206, 207, 249, 250,265, 266 Indictor, N., 300, 322 Ingberman, A. K., 272,317 Ingold, K. U., 308, 321 Irving, H., 222, 266 Ishidate, H., 220, 221, 266 Ishigami, T., 55, 61 Ishikawa, K., 249, 269 Ishikawa, R. M., 10, 61 Itatani, Y., 259, 269 Ittel, S. D., 113, 118, 159, 162 Iwatsubo, M., 182, 264 Iyer, P. S., 74, 151, 160 Jackman, L. M., 73, 159 Jackson, 8 . G., 187, 190, 206, 267 Jackson, D., 16, 61 Jackson, G. E., 220, 221, 264 Jackson, G. L., 21 1, 262 Jackson, P. F., 191, 263 Jaffe, M. H., 64, 146, 161 James, M. N. G., 187, 188, 266 Jameson, C. J., 71, 159 Janowicz, A. H., 50, 60 Janzen, E. G., 274, 275, 315, 321 Jarret, R. M., 81, 161 Jaurin, B., 252, 266 Jawad, J. K., 15,43,60 Jawdosiuk, M., 278, 280, 284, 285, 291, 292, 295, 296, 318, 321 Jayaraman, B., 286, 321 Jayaraman, H., 195, 269 Jeffery, J. C., 36, 58 Jeffrey, G. A,, 186, 266 Jensen, F. R., 19, 21, 23, 27, 60, 64, 102, 107, 159 Jencks, W. P., 182, 218, 221, 223, 234, 235, 238, 245, 248, 249, 253, 254, 262, 265, 266, 268 Jeng, S., 189, 197, 267 Jennings, K. R., 258, 261, 262

330

JeremiC, D., 104, 105, 159, 162 Jesson, J. P., 113, 159 Jeuell. C. L., 141, 160 Jiang, W., 306, 321 Jimenez, P., 16, 17, 59 Johnson, C. A., 64,71, 161 Johnson, D. A., 259, 266 Johnson, K., 180, 264 Johnson, M. D., 19, 30, 58, 59, 60 Johnson, R. W., 19, 20, 60 Johnston, D. B. R., 259, 264 Jones, W. D., 52,60, 297, 298, 318 Jonsall, G., 64,93, 94, 95, 119, 158 Jonsson, L., 308, 31 7 Jordan, R. F., 12, 60 Joris, B., 177, 180, 182, 252, 264, 265 Joseph-Natan, P., 72, 159 Jugelt, W., 309, 318 Jung, S., 258, 263 Jurgensen, G., 214, 264 Kagan, H. B., 197, 266 Kahl, A. A., 300, 318 Kahle, A. D., 84, 161 Kaji, A., 277, 300, 304, 320 Kake, Y., 314, 316 Kalinowski, H.-O., 64,70, 93, 96, 97, 128, 137, 158, 159, 160 Kamirnura, A., 300, 304, 320 Kanao, S.. 220, 221, 266 Kanayama, K., 215, 270 Kang, J., 10, 59 Kapp, D. L., 310, 314,319 Karle, J., 82, 160 Kates, M. R., 64,72, 73, 75, 79, 80, 124, 125, 128, 131, 133, 148, 149, 150, 151, 158, 160, 161 Kaupp, G.. 295,320 Kehlen, H., 192, 268 Kehlen, M., 192, 268 Kehoe, D. C . , 294, 316 Kelly, D. P., 141, 160 Kelly, J. A., 180, 182, 184, 265, 266 Kelm, H.. 49, 62 Kennedy, J. D.. 28, 61 Kerber, R. C., 276, 280, 318 Kern, M.. 233, 268 Kershner, L., 195, 269 Kessler, D. P., 209, 210, 212, 213, 258, 262, 266

AUTHOR INDEX

Kessler, H., 194, 266 Kestner, M. M., 277, 319 Kezdy, F., 199, 21 I , 262 Khanna, R. K., 285, 286, 287, 288, 289, 295, 297, 298, 303, 307,320 Kharasch, M. S., 272, 273, 318 Khosla, S., 252, 264 Khurana. J. M., 297, 298, 299, 321 Kiefer. G. W., 10, 61 Kill, R. G., 300, 318 Kim, J. K., 276, 318 Kim, M., 192, 268 Kimura, T., 227, 270 King, R. B., 2. 12, 14, 44, 60 King, T. J., 240, 265 Kinney, R. J., 297, 298, 318 Kirby, A. J., 241, 249, 266 Kirby. S. M., 202, 268 Kirchen. R. P., 119, 120, 122, 138, 162 Kitching, W.. 27, 28, 61 Kiya, E., 201, 214, 236, 248, 253, 254, 2 70 Klein, D.. 182, 184, 207, 265, 266 Kleinman, L. I., 66, 163 Knevel, A. M., 209, 210, 212, 213, 258, 262, 266 Knickel, B., 23, 60 Knitter, B., 215, 268 Knott-Hunziker, V., 252, 266 Knowles, J. R., 252, 264 Knox, J. R., 180, 266 Koch, E.-W., 136, 137, 160, 162 Kochi, J. K., 4, 43 61, 62, 272, 294, 306, 318, 320 Kocy. O., 240, 270 Koehl, W. J., 308, 318 Koelsch, C. F., 272, 318 Koermer, G. S., 27, 28, 61 Kokesh, F. C., 254, 265 Kolb, V. M., 285, 319 Kolfini, J. E., 207, 261 Komiyama, M., 232, 266 Komoto, R. G., 54, 59 Konishi, M., 26, 60 Kopelevich, M., 106, 158 Koppel. G. A,, 206, 265 Kornblum, N., 276, 277, 278, 279, 280, 282, 283, 284, 285, 296, 297, 298, 301, 318 Korzan, D. G., 297, 319

AUTHOR INDEX

Koshiro, A., 215, 265 Kossi, R., 309, 319 Kostenbauder, H. B., 224, 265 Kozima, S., 28, 62 Krabbenhoft, H. A., 194, 266 Kramer, A. V., 23, 61 Krane, J., 104, 158 Kreevoy, M. M., 300, 320 Kresge, A. J., 72, 160, 210, 266 Krishnamurthy, V. V., 74, 151, 160 Kristiansen, H., 214, 264 Kruchten, E. M. G. A., van 134, 159 Kruger, J. D., 125, 148, 160 Kubota, M., 10, 36, 59, 61 Kudryavtsev, L. F., 275, 320 Kuivila, H. G., 27, 28, 61 Kukovitskii, D. M., 299, 322 Kumada, M., 26, 60 Kump, R. L., 16, 17, 59 Kunzer, H., 72, 158 Kurosawa, H., 300, 319 Kutal, C., 312, 316, 322 Labinger, J . A., 16, 23, 24, 59, 61 Labischinski, H., 174, 268 Laht, A., 103, 160 Lamanna, W., 117, 158 Lamb, R. C., 275,319 Lamotte-Brasseur, J., 177, 265 Landvatter, E. F., 56, 61 Lang, R. W., 65, I59 Lappert, M . F., 4, 61 Larsen, C., 233, 253, 263, 266 Lau, K. S. Y., 4, 24, 25, 32, 61, 62 Laue, H. A. H., 294,317 Laungani, D., 77, 82, 129, 159 Laurent, G., 182, 207, 266 Lavagnino, E. R., 190, 206, 267 Lavin, M., 53, 59 Lebouc, A., 309,317, Lechevallier, A., 278, 281, 316 LeClerc, G., 313, 316 Led, J. J., 64, 160 Lednor, P. W., 4, 61 Ledwith, A., 310,317,319 Lee, B. J., 91, 159 Lee, C . L., 39, 58 Lee, S.-L., 191, 270 Lee, W. S., 259, 270

331

Le Fevre, R. J . W., 194, 262 Leininger, H., 89, 158 Leipert, T. K., 82, 160 Lempert, K., 297, 298,321 Levashov, A. V., 226, 267 Levi, B. A., 68, 159 Levine, B. B., 233, 266 Levy, G. C., 189, 266 Lewis, G , J., 30, 59 Leyendecker, F., 108, 158 Leyh-Bouille, M., 177, 180, 184, 264, 265 Lichter, R. L., 189, 190, 206. 268 Liler, M., 210, 266 Limbach, H. H., 125, 148, 160 Limbert, M., 207, 269 Lin, L., 82, 158 Linder, P. W., 220, 264 Lippmaa, E., 103, 160 Lloyd, J. R., 47, 129, 160 Loffet, A., 184, 264 Loffler, K., 308, 319 Lok, S. M., 125, 148, 160 Lokaj, J., 300, 318 Long, K. M., 41,61 Longridge, J. L., 209, 251, 262, 266 Lopez, J., 314, 319 Loukas, S. L., 209, 213, 263 Louw, W. J., 3 , 6 1 Lowrey, A. H., 82, 160 Lu, J.-J., 293, 319 Lucchi, C., 180, 264 Luche, J. L., 197, 266 Lukovkin, G. M., 82, 161 Lumbreras, J. M., 215, 216, 266 Lund, F., 259, 262 Lundin, A. F., 313, 319 Lunn, W. H. W., 202,206,262 Lutz, A., 207, 262, 269 LyEka, A., 84, 159 Ma, K. W., 297, 298, 318 MacConnell M. M., 72, 159 Macho, V., 125, 126, 148, 160 Maciejewicz, N. S.,206, 259, 264 MacLaury, M . R., 14, 37,59 Madan, V., 19, 21,60 Magnus, P. D., 313,316 Mahe, C., 294, 317

332

Maier, G., 97, 160 Maitlis, P. M., 39, 48, 60 Majerski, Z., 72, 160 Majima, T., 312, 319 Mak, S., 274, 275, 321 Makosza, M., 284, 295, 296,318, 321 Maldonada, J., 281, 317 Malhotra, S. K., 313, 319 Malihowski, E. R., 189, 267 MalojEit, R., 87, 162 Manhas, M. S., 189, 197, 267 Mania, D., 259, 266 Manthey, J. W., 280, 318 Marchand-Brynaert, J., 240, 268 Mares, F., 310, 313, 321 Marquet, A., 180, 264 Marsh, M. M., 202, 262 Marshall, A. C., 258, 261, 262 Marshall, D. R., 195, 264 Marten, D. F., 45, 59 Martin, A. F., 233, 241, 243, 244, 267 Martin, J. C., 193, 194, 267, 268 Martinek, K., 223, 225, 226, 267 Marzilli, L. G., 10, 62 Maslen, E. N., 187, 188, 261 Masters, A. F., 57, 61 Matisons, J. G . , 294, 316 Matschiner, H., 192, 268 Matsuda, M., 224, 269 Mattes, S. L., 31 1, 319 Matthews, W. S., 285, 319 Mayake, H., 279,319 Mayer, M. G., 66, 158 Mayeste, R. J., 220, 264 McAteer, C. H., 116, 160 McCleverty, J., 32, 61 McLaren, K. L., 125, 129, 132, 163 McLellan, D., 197, 221, 255, 258, 261, 265 McLelland, R. A., 210, 267 McMillian, F. L., 300, 318 McMullen, C. H., 272, 317 McMullen, R. K., 186, 266 McMurry, J. E., 120, 160 Meesschaert, B., 184, 261 Meijs, G. F., 284, 286, 319 Meintzer, C., 308, 321 Meinwald, J., 95, 158 Melander, L., 65, 66, 69, 160 Menger, F. M., 225, 267

AUTHOR INDEX

Menzies, I. D., 313, 316 Merz, Jr., K. M. 129, 159 Metelco, B., 72, 160 Meyer, W. P., 194, 267 Meyers, C. Y., 285,319 Michael, R. E., 276, 280, 318 Mihelcic, J. M., 50, 59 Miles, W. H., 112, 158 Milhailovit, M. L., 104, 105, 159, 162 Miller, I. M., 166, 267 Miller, L. L., 286, 318 Miller, M. A., 197, 261 Milne, C. R. C., 40, 62 Milosavlievit, S., 104, 105, 159, 162 Milstein, D., 31, 38, 55, 56, 61 Mincy, J. W., 209, 212, 262 Minhas, H. S., 201, 267 Minisci, F., 273, 302, 308, 31 7, 319 Mioduski, J., 95, 158 Mislow, K., 192, 269 Mitsudo, T., 54, 62 Mitton, C. G., 255, 267 Miyake, H., 300, 304,320 Miyarnoto, E., 201, 214, 224, 236, 248, 253, 254, 269, 270 Mizukarni, Y., 198, 212, 214, 215, 270 Modro, T. A,, 210, 267, 270 Moews, P. C., 180, 266 Moll, F., 198, 267 Mollison, G. S. M., 191, 192, 265 Molyneux, P.,232, 267 Mondelli, R., 189, 206, 267 Monse, E. U., 69, 162 Moore, P., 116, 160 Moore, S. S., 52, 60 Moreau, C., 189, 206, 264 Morehouse, S. M., 53, 59 Moreno, R., 180, 264 Morgan, K. J., 191, 263 Morin, R. B., 186, 187, 190, 201, 206, 267, 269 Morris, A., 202, 268 Morris, G. E., 116, 160 Morris, J. J., 195, 197, 201, 221, 233, 234, 236, 237, 238, 239, 241, 242, 243, 244,245, 248, 255, 258, 261, 265, 267 Morris, R. B., 187, 267 Mortimer, C. T., 50, 61 Mosher, M. W., 308, 321

AUTHOR INDEX

Moskau, D., 72, 159 Motevalli, M., 47, 61 Motokawa, K., 190, 206, 267 Moye, A. J., 274, 275, 321 Mudryk, B., 278, 280, 285, 291, 292, 318, 321 Mueller, R. A., 190, 206, 267 Muir, W. R., 22, 24, 34, 61 Murakami, K., 190, 206, 267 Muranishi, S., 227, 270 Mureinik, R. J., 33, 61 Murray, H. H., 85, 161 Musser, M. T., 280, 318 Myhre, P. C., 125, 126, 129, 132, 148, 160, 163

333

Nitay, M.. 16, 61 Noack, K., 19,59 Nonhebel, D. C., 272,319 Noordik, J . H . , 41, 60 Norman, R. 0. C., 294, 317 Norris, R. K., 277, 279, 280, 281, 282, 284, 285, 316, 31 7, 319, 321 Norton, J. R., 12, 60 Norvilas, T. T., 190, 203, 206, 249, 250, 266 Noyd, D. A., 284, 316 Noyori, R.. 55, 61 Nudenberg, W., 272, 318 Nugent, W. A,, 306, 320

Oakenfull, D. G., 218, 249, 253, 254. 268 Nagarajan, R., 166, 190, 263, 265, 267 O’Csllaghan. C. H. O., 202, 268 Nagata, W., 206, 267 Ochiai, E. I., 41, 61 Nakajima, T., 136, 160 O’Connor, C. J., 210, 212, 231, 265. Nakamura, C., 254, 266 268 Nakano, O., 215, 270 Ohno, A., 300, 320 Nakashima, E., 249, 269 Ohtani. M., 206, 207, 267 Nakasone, A,, 312, 319, 320 Ohtani. Y., 33, 6 1 Nakatuska, T., 277, 320 Oka, S., 300, 320 Narisada, M., 206, 207, 267 Okada, H., 300,319 Nash, C. P., 194, 262 Okajima, T., 54, 62 Nayler, J . H. C., 198, 261, 263, 264 Okazawa, N., 119, 122, 162 Oki, M., 73, 160 Neese, R. A., 102, 159 Olah, G. A., 74, 123, 133, 136, 141, Nelsen, S. F., 310, 314, 319 146, 151, 160 Nelson, H . D., 232, 267 Oliver, A, J., 32, 61 Nelson, G. L., 189, 266 Olmstead, M. M., 39, 58 Neunteufel, R. A., 31 1, 319 Olson, J. M., 65, 159 Neugebauer, W., 88, I61 Olszowy, H., 27, 28, 61 Newcombe, P. J., 281,319 Ono, N., 277, 300, 304,319, 320 Newton, €3. N., 279, 319 Onoue, H., 190, 206, 207, 267, 268 Newton, G. G. F., 166, 267 Ordonez. D., 215, 216, 266 Ngoviwatchai, P., 304, 306, 321 Orpen, A. G., 112, 162 Nguyen-Distbche, M., 177, 265 Osborn. J. A,, 16, 23, 24, 59.61 Nicholson, B. K., 294, 316 Niebergall, P. J., 218, 220, 263 Osipov. A. P., 223, 226, 267 Osten. H.-J., 71. 159 Nieto, M., 180, 264 Osterman, V. M., 72, 159 Nigham, A., 297, 298, 299, 321 Nikaido, H., 174, 267 Ostovic, D., 300, 320 Ovary, Z., 233, 266 NikoletiC, M., 137, 162 Nishide, K., 249, 269 Owens. K., 285, 321 Nishikawa, J., 190, 205, 206, 207, 267, 268, 269 Nishimura, K., 224, 269 Pabon. R. A,. 31 I , 316. 320

AUTHOR INDEX

334

Pac, C., 312, 319, 320 Page, J. E., 187, 265 Page, M. I., 182, 185, 193, 195, 196, 197, 199, 201, 202, 203, 205, 206, 207, 209, 210, 211, 212, 213, 214, 215, 218, 219, 220, 221, 222, 223, 224, 225, 227, 228, 229, 230, 232, 233, 234, 235, 236, 237, 238, 239, 240. 241, 242, 243, 244, 245, 246, 247, 248, 249, 250, 25 1, 252, 254, 255, 256, 258, 259, 261, 261, 262, 263, 265, 266, 267, 268 Pagnotta, M., 193, 264 Palacios, S. M., 286, 320 Palazzi, A,, 49, 62 Pan, Y., 74, 148, 159 Panek, E. J., 277, 279, 282, 321 Pannell, K. H., 16, 61 Panossian, R., 221, 262 Panoy, T. E., 285, 319 Park, W. S., 284, 286, 316 Parker, C. W., 233, 268 Parker, W. L., 168, 269 Parthasarathy, R., 197, 266, 268 Parshall, G., 51, 61 Paschal, J. W., 189, 190, 206, 268 Pasini, A,, 1 1 , 15, 16, 44, 48, 62 Patel, G. S., 206, 264 Patin, H., 294, 317 Paukstelis, J. V., 192, 268 Pauson, P., 272, 318 Pawelczyk, E., 21 5, 268 Pawliczek, J.-B., 93, I59 Pearson, R. G., 10, 12, 14, 16, 19, 20, 22, 24, 31, 33, 34, 35, 48, 60, 61 Pecoraro, J. M., 280, 292, 321 Pehk, T., 103, 160 Pelech, B., 96, 158 Penasse, L., 207, 262, 269 Penenory, A. B., 285, 320 Perkins, H. R.. 180, 184, 264 Perkins, I., 12, 59 Perland, R. A,, 53, 61 Perron, Y. G., 187, 201, 270 Perronet, J., 207, 262 Petersen, S. B., 64, 160 Petrongolo, C., 206, 268 Pettit, R., 54, 59 Petukhov, G. G., 275, 320 Petursson, S., 252, 266

Pfaendler, H. R., 190, 191, 201, 268 Pfeiffer, R. R., 190, 203, 206, 249, 250, 266 Pflug, G. R., 253, 269 Pfluger, F., 33, 43, 48,60 Pham, T. N., 297, 298,299,316 Pickard, A. L., 33, 58 Pierini, A. B., 285, 320 Pierpont, C., 41, 60 Pierson, C., 308, 317 Pinnick, H. W., 277, 280, 296, 318, 319 Pitt, G. J., 187, 268 Plackett, J. D., 196, 233, 262 Poiil, R. L., 12, 14, 60 Poli, R., 47, 61 Poortere, M. De, 240, 268 Pople, J. A., 186, 193, 265, 266 Popov, V. I., 285, 320 Porta, O., 273, 31 7 Porter, R. D., 141, 160 Portnoy, C. E., 225, 267 Potter, A,, 305, 321 Poulos, A. T., 31, 35, 61 Powell, M. F., 72, 160 Pracejus, H., 192, 268 Pragst, F., 309, 318 Prakash, G. K. S., 74, 123, 133, 136, 141, 146, 151, 160 Pratt, R. F., 201, 250, 252, 262, 263, 264, 268 Pratt, W., 64, 131, 161 Prescatori, E., 206, 268 Presti, D. E., 202, 262 Pribula, C. D., 16, 61 Proctor, P., 193, 195, 196, 197, 199, 201, 202, 203, 205, 206, 207, 209, 210, 211, 212, 213, 214, 215, 218, 219, 220, 221, 222, 223, 240, 241, 249, 250, 251, 252, 255, 258, 261, 261, 263, 265, 268 Pross, A., 315, 320 Prousek, J., 292, 295, 320 Puar, M. S., 207, 261 Puddephatt, R. J., 4, 15, 39, 40, 43, 50, 58, 60, 61

Quinn, C. B., 194, 266 Quirk, J. M., 50, 59

AUTHOR INDEX

Radom, L., 192. 193, 194, 262, 263, 265, 268 Rajaram, J., 33, 61 Rakhomankulov, D. L., 299, 322 Rakshit, D., 277, 278, 279, 316 Ramasami, T., 42, 62 Randahawa, G., 197, 221, 255, 258, 261, 265 Randles, D., 280, 319 Ranganayakulu, K., 119, 120, 122, 126, 128, 129, 130, 162 Ranghino, G., 206, 268 Rao, V. S. R., 192, 270 Rappoport, Z., 75, 160 Rasmussen, J. R., 20, 59 Rastrup-Andersen, N., 259, 262 Ratcliffe, R. W., 259, 264 Rauchfuss, T. B., 56, 61 Rauk, A,, 119, 120, 122, 162 Rawdah, T. N., 104, 158 Razuvaev, G. A., 275,316, 320 Ree, B., 193, 268 Reichenbach, G., 48, 58 Reinicke, B., 174, 268 Requena, Y., 184,264 Reuben, J., 65, 74, 160 Reuver, J. F., 72, 162 Reynolds, C. H., 145, 159 Reynolds, W. F., 210, 267 Rhodes, C. T., 232, 267 Richards, E., 202, 250, 265 Richardson, F. S., 191, 262, 268 Richardson, G. D., 277, 316 Rick, E. A., 42, 43, 60 Rico, I., 285, 301, 320 Rieger, P. H., 294, 316 Riggs, N. V., 192, 263, 268 Rihs, G., 190, 191, 201, 268 Ritchie, G. L. D., 194, 262 Roberts, G. C. K., 182, 209, 213, 263 Roberts, J. D., 141, 142, 144, 146, 160, 189, 270 Roberts, M. K. S., 313, 318 Roberts, R. M. G., 300, 320 Robertson, G. B., 36, 58 Robinson, I. R., 248, 263 Robinson, L., 225, 263 Robinson, M. J. T., 74, 82, 100, 101, 102, 160, 161 Robinson, R., 166, 261

335

Roets, E., 191, 263 Rogers-Low, B. W., 166, 263 Romsted, L. S., 226, 268 Roper, W. P., 7, 59 Ros, F., 277, 280, 285, 291, 292, 321 Rosen, K. M., 82, 161 Rosenblum, M., 16, 61 Rousseau, H., 278, 281, 316 Rossi, R. A,, 285, 286, 320 Roth. H. D., 314, 320 Ruble, J. R., 186, 266 Russell, G. A., 274, 275, 276, 277, 278, 279, 280, 282, 284. 285, 287, 288. 289, 290, 291, 292, 293, 295, 296, 303, 304, 305, 306, 307, 315, 318, 320, 321 Sackett, J. R., 16, 17, 59, 110 Saillard. J. Y., 3, 53, 62 Sakaguchi, T., 220, 221, 266 Sakurai. H., 312, 319, 320 Sands, T. H., 166, 265 Sandstrom, J., 73. 161 San Filippo, Jr. I., 28, 62 Santiago, A. N., 285, 286, 320 Saritillan, R. L., 72, 159 Sardella, D. J., 72, 159 Sato, F., 227, 270 Saunders, M., 64, 69, 71, 72, 73, 75, 80, 81, 82, 85, 98, 102, 121, 123, 124, 125, 128, 131, 133, 142, 143, 146, 149, 150, 151, 155, 158, 161 Saunders, S., 64, 71, 161 Saunders, Jr., W. H., 65, 69, 160 Savelli, G., 223, 263 Sawyer, J. F., 40, 62 Scanlon, W. B., 190, 206, 267 Schaad, L. J., 72, 162 Schaefer, W. P., 91, 159 Schanck, A., 189, 203, 206,263, 264, 268, 269 Schelechow, N., 259, 264 Schenck, T. G., 40, 62 Schilling, M. L. M., 314, 320 Schleyer, P. v. R., 64, 75, 88, 92, 128, 135, 158, 161, 193, 263 Schlosser, M., 88, 161 Schmitt, P., 83, 84, 159 Schmitz, L. R., 131, 132, 162

336

Schneider, C. H., 233, 253, 269 Schneider. J., 137, 161 Schowen, R. L., 195, 255, 267, 269 Schrauzer, G. N., 8, 12, 14, 41, 42, 62 Schrinner, E., 207, 269 Schultz, A. J., 109, 118, 161, 162 Schwartz. M. A., 209, 212, 215, 233, 241, 242, 248, 253, 269 Schwenk, E., 300,318 Scordamaglia, S., 206, 268 Scott, J. D., 40, 61 Sears, Jr., C. T. 36, 59 Sen, L. A., 50, 62 Senoff, C. V., 10,60 Servis, K. L., 72, 161 Sesana, G., 273, 317 Sessions, R. B., 310, 316 Seybold, G., 131, 161 Sezaki, H., 227, 270 Shapet’ko, N. N., 82, I61 Shapiro, J., 233, 268 Shapiro, M. J., 84, 161 Shapley, J. R., 109, 11 I , 161, 162 Sharma, H. N., 221, 263 Shaskus, J. J., 201, 268 Shaw, B. L., 10. 34, 35, 36, 59,60, 62 Shchupak, G. N., 285, 316 Shealer, S. E.. 311, 317 Shechter, H., 295, 322 Shimizu, T., 249, 269 Shiner, Jr.. V. J., 68, 70, 92, 137, 152, 162 Shio, T., 300, 320 Shriver, D. F., 12, 62 Shudo, L., 194, 265 Shue, F.-F., 72, 161 Siddall, 111, T. H., 194, 269 Sidot, C.. 286, 316 Siehl, H.-U., 121, 136, 138, 139, 142, 143, 146, 153, 155, 161, 162 Silbermann, J., 28, 62 Simig, G., 297, 298, 321 Simmons, W., 314, 317 Simon, G. L., 186, 187, 190, 201, 269 Simonet, J., 309, 317, 319 Singh, B. P., 119, 120, I62 Singh, H. K.. 279, 298, 300, 304, 319, 32 I Singh, P. R., 286, 297, 299, 321 Sisido, K., 28, 62

AUTHOR INDEX

Sklavounos, C. G., 258, 263 Slack, D. A., 19, 62 Slusarchyk, W. A., 207, 261 Smentowski, F., 275,321 Smith, C. A., 48, 60 Smith, C. R., 210, 269 Smith, G. B., 250, 265 Smith, K., 272, 317 Smith, L. A., 64, 107, I59 Smith, R. G., 277, 279, 280, 282, 296, 297, 298, 319 Smith, S. A., 182, 265 Snow, D. H., 280, 318 S G ~ O ~ V. O VI.,, 26, 62 Sommer, J., 141, I60 Sorensen, T. S., 119, 120, 122, 125, 126, 128, 129, 130, 131, 132, 138, I62 Sosnovsky, G., 273, 318 Southgate, R., 261, 263 Sowinski, A. F., 52, 60 Spanswick, J., 308, 321 Sperati, C. R., 41, 61 Spindel, W., 69, I62 Spitzer, W. A., 202, 206, 262, 265 Spratt, B. G., 174, 180, 269 Squillacote, M. E., 142, 144, 146, 160 Srinivasan, P. R., 189, 190, 206, 262, 268 Stackhouse, J., 192, 269 Stainbank, R. E., 36, 62 Stam, M., 225, 263 Stapley, E. O., 166, 267 Staral, J. S., 141, 146, 160 Stark, W. M., 166, 267 Steiger, H., 49, 62 Stein, A. R., 300, 321 Steinhoff, G., 193, 270 Stern, M., 69, I62 Stevens, J. B., 210, 270 Stewart, W. E., 194, 269 Stille, J. K., 4, 24, 25, 26, 32, 58, 61, 62 Stirling, C. J. M., 195, 203, 264, 269 Stivers, E. C., 72, I62 Stobart, S. R., 41, 59 Stone, G. A., 112, 162 Stothers, J. B., 189, 264 Stove, E. R., 261, 263 Strahle, M., 88, 161 Straniforth, S. E., 187, 265 Straus, D. A., 91, 159

AUTHOR INDEX

Strege, P. E., 24, 62 Strom, E. T., 274, 275, 315,321 Strominger, J. L., 177, 180, 181, 184, 256, 269, 270 Stryker, J. M., 53, 62 Stuchal, F. W., 279, 280, 282, 301, 318, 319 Stucky, G. D., 109, 118, 161, 162 SU, W.-Y., 284, 286, 316 Sugita, E. T., 218, 220, 263 Suib, S. L., 109, 161 Sun, J. Y., 10, 59 Sundberg, J. E., 285, 31 7 Sunko, D. E., 87, 137, 146, 152, 155, 157, 158, 162 Surh, Y. S., 201, 268 Surzur, J.-M., 281, 317 Sutcliffe, L. H., 309, 316 Swain, C. G., 72, 162 Swarbick, J., 232, 267 Sweet, R. M., 186, 187, 188, 190, 202, 206, 269 Swiger, R. T., 280, 319 Sykes, R. B., 168, 269 Symons, M. C. R., 277, 279,316 Szele, I., 155, 157, 162 Szepsey, P., 297, 298, 321

Thelan, P. J., 308, 322 Thomas, R. J., 195, 262 Thompson, D. H. P., 286. 318 Thorne, D. L., 56, 62 Timms, D., 209, 266 Tin, K.-C., 191, 270 Tipper, D. J., 175. 177, 180, 269 Titchmarsh. D. M., 30, 59 Tiwari, A,. 221, 263 Tiwari, C . P.. 221, 263 Todres. Z . V., 299, 322 Toeplitz. B., 207. 261 Tomasz. A.. 174, 178, 269 Tomida. H., 223. 248. 269 Tonellato. V., 231, 262 Toney. M . K., 275,319 Toppet, S., 258, 263 Tori, K., 190, 205. 206, 207. 267, 268, 26Y Toscano, P. J., 10, 62 Toth. G.. 297, 298, 321 Towner, R. D., 206, 265 Traylor. T. G., 27. 28, 61 Tribble, M . T., 197, 261 Trimerie. B., 240, 268 Trost. B. M., 24, 26, 53. 62 Troxell, T. C., 191, 268 Truce, W. E., 293. 319 Tsou, T. T., 43, 62 Taguchi, K., 220, 221, 266, 269 Tsoucaris, G., 197. 266 Takasuka, M., 190, 206, 207, 267, 268, Tsuji, A.. 198, 201. 202, 207, 212. 214, 269 2 15, 224. 236, 248, 249, 253. 254, Takata, R., 314,316 259, 269, 270 Takayama, H., 299, 316, 317 Turetzky, M. N.. 299. 317 Takegami, Y., 54,62 Turner-Jones, A,, 166, 263 Takizawa, K., 28, 62 Tutt. D. E., 233, 253. 269 Tallec, A,, 309, 317 Tamas, J., 297, 298, 321 Tamura, R., 277, 300, 304,320 Ugo, R.. I I , 15, 16, 44, 48, 62 Tan, A.-L., 231, 268 Uguagliati, P., 49. 62 Tanaka, H., 261, 265 Ulsen. J., 93, 159 Tanaka, M., 54, 62 Umeda. I., 55. 61 Tang, R., 310, 313, 321 Tang, Y. S., 72, 160 Tanner, D. D., 297, 298, 300, 304, 305, Vali, Z.. 297, 298, 321 308,316, 321 Valmas. M. D., 277. 278. 279, 316 Tashtoush, H., 277, 304, 306, 321 Van-Catledge, F. A,, 1 1 3 , 159 Teasley, M. F., 310, 314, 319 Vandekerkhove. J., 182, 265 Telkowski, L., 64, 125, 138, 149, 150, Vanderhaeghe, H.. 182, 184, 191, 258, 151, 158, 161, I62 261, 263, 264, 265

337

338

Van Duong, K. N., 31, 62 Vanelle, P., 278, 280, 281, 317 Van Meerssche, M., 189, 203, 206, 263, 264, 268, 269 Van Scoy, R. M., 300, 318 Vaska, L., 2, 3, 10, 62 Vastine, F., 10, 59 Ventura. P., 189, 206, 267 Verhoeven, T. R., 26, 53, 62 Vieth, H. M., 125, 148, 160 Vijayan. K.. 188, 269 Viout. P., 231, 265 Vishveshwara, S., 192, 270 Vismara, E., 302, 319 Vogel. P., 64, 69. 131, 146. 149, 161, 162 Waegell. B., 297, 318 Wagstaff, K., 119, 162 Wakselman, C., 285, 301, 320 Waley, S. G., 252, 266 Walker, G. E., 75, 150, 161 Waller, E. R., 202, 268 Walling, C., 123, 126, 162, 275, 300, 322 Walter, H., 79, 121, 153, 155, 162 Wan, P., 210, 270 Watanabe, Y., 54, 62 Waters, W. A,, 272, 319, 322 Watson, P. L., 17, 62 Waugh. J.. 27, 28, 61 Wawzonek, S.. 308,322 Wax, J., 53. 62 Waxman, D. J., 180, 181, 256, 270 Wayner, D. D. M., 314, 322 Weber, J. H.. 12, 62 Webster. P., 240, 268 Wehrli, F. W., 102. 104, 162 Wei, C.-C., 202. 270 Weigele, M., 202, 270 Weiner, S. A,. 275. 321 Weiss, A,. 220. 270 Weiss, K.. 272, 317 Weitzberg, H., 42, 59 Weitzberg, M., 33, 61 Wells. J. S., 168, 269 Wenderoth, B., 297, 298, 299, 316 Wennerstrom, H., 71, 82, 129, 159 Wesener, J. R., 72, 159

AUTHOR INDEX

Weuste, B., 70, 93, 96, 137, 158 Whalen, R., 308, 31 7 Wheeler, W. J., 190, 203, 206, 249, 250, 266 Wheland, G. W., 193, 270 Whipple, E. B., 104, 162 Whitesides, G. M., 19, 20, 21, 27, 52, 58, 59, 60, 62 Whitney, J. G., 166, 267 Wiberg, K. B., 64, 131, 161 Widdowson, D. A., 300, 318 Widmer, J., 279, 283, 297, 298, 319 Wilham, W. L., 190, 203, 206, 207, 249, 250, 265, 266 Willi, A. V., 65, 162 Williams, A., 210, 249, 270 Williams, I. H., 106, 162 Williams, J. M., 109, 118, 161, 162 Williams, M. L., 294, 316 Williams, R. J. P., 222, 266 Williamson, K. L., 189, 270 Wilkinson, G., 5 , 14, 32, 42, 47, 59, 60, 61 Wilkinson, M . P., 50, 61 Wilmot, I. D., 313, 318 Windgassen, R. J., 12, 62 Winter, M. J., 45, 58 Winter, S. R., 54, 59 Wirth, D. D., 31 I , 316 Wirthlin, T., 102, 162 Wiseman, J. R., 194, 266 Wittig, G., 193, 270 Wojtkowski, P. W., 240, 270 Wolf, J. F., 310, 313, 321 Wolfe, S., 187, 191, 201, 259, 270 Wolfsberg, M., 66, 71, 72, 162, 163 Wong, N., 125,162 Wong, P. C., 314, 322 Wong, P. K., 25, 61, 62 Woodward, R. B., 184, 185, 187, 190, 191, 201, 268, 270 Woolfenden, S. K., 284, 285, 317 Workman, J. D. B., 82, 161 Worsfold, D. J., 88, 158 Wright, D. E., 248, 263 Wrobel, Z., 285, 291, 292, 321 WU, G.-M., 242, 269 Wullbrandt, D., 72, 159 Wyckoff, J. C., 308, 317 Wynand, J. L., 38, 60

339

AUTHOR INDEX

Yagupol’skii, L. M., 285, 316, 320 Yamagishi, A., 33, 61 Yamamoto, H., 300, 320 Yamamoto, K., 54, 62 Yamamura, K., 279, 319 Yamana, T., 198, 201, 202, 207, 212, 214, 215, 224, 236, 248, 249, 253, 254, 259, 269, 270 Yang, D., 297,298, 300, 304,321 Yang, G. K., 30,62 Yang, J.-R.,72, 159 Yannoni, C. S., 125, 126, 129, 132, 148, 160, 163 Yano, S.,72, 163 Yashiro, M., 72, 163 Yasuda, M., 300, 319 Yasufuka, K., 312,322 Yates, K., 210, 267, 269, 270 Yatsimirski, A. K., 223, 225, 226, 267 Yatsuhara, M., 227, 270

Yavari, A., 39, 58 Yavari, I., 141, 146, 160 Yeh, C.-Y., 191, 262,268 Yue, H. J., 310, 313, 321 Yoneda, G., 49,62 Yoshida, T., 190, 206, 207, 267 Yoshikawa, S., 72, 163 Zajac, M., 215, 268 Zeilstra, J. J., 277, 322 Zeimer, E. H. K., 50, 62 Zeldrin, L., 295, 322 Zeliver, C., 197, 266 Zhil’tsov, S. F., 275, 316, 320 Ziegler, T., 12, 62 Zlotskii, S. S., 299, 322 Zorin, V. V., 299, 322 Zuanic, M., 72, 160 Zugara, A., 221, 270

Cumulative Index of Authors Ahlberg. P., 19, 223 Albery, W. J., 16, 87 Allinger, N. 1.. 13, 1 Anbar, M.. 7, 115 Arnett, E. M., 13, 83 Bard, A. J., 13, 155 Bell, R. P., 4, I Bennett, J. E., 8, 1 Bentley, T. W., 8, 151; 14, 1

Berger, S.. 16, 239 Bethell, D., 7, 153; 10, 53 Blandamer, M. J.. 14, 203 Brand, J. C. D., 1, 365 Brandstrom, A,, 15, 267 Brinkman, M. R., 10, 53 Brown, H. C., 1, 35 Buncel, E., 14, 133 Bunton, C. A,, 22, 213 Cabell-Whiting, P. W., 10. 129 Cacace, F., 8, 79 Capon, B., 21, 37 Carter, R. E., 10, I Collins. C. J.. 2, 1 Cornelisse, J., 11, 225 Crampton. M. R., 7, 21 1 Davidson, R. S., 19, I ; 20, 191 Desvergne, J. P., 15, 63 de Gunst, G . P., 11, 225 de Jong. F., 17, 279 Dosunmu, M. I., 21, 37 Eberson, L.. 12, 1; 18, 79 Engdahl, C., 19, 223 Farnum, D. G., 11, 123 Fendler, E. J., 8, 271 Fendler, J. H., 8, 271; 13, 279 Ferguson, G., 1, 203 Fields, E. K., 6, 1 Fife, T. H.. 11, 1

Fleischmann, M., 10, 155 Frey, H. M., 4, 147 Gilbert, B. C., 5, 53 Gillespie, R. J., 9, 1 Gold, V., 7,259 Goodin, J. W., 20, 191 Gould, I. R., 20, 1 Greenwood, H. H., 4, 73 Hammerich, O., 20, 55 Havinga, E., 11, 225 Henderson, R. A., 23, 1 Henderson, S., 23, 1 Hibbert, F., 22, 113 Hine. J., 15, 1 Hogen-Esch, T. E., 15, I53 Hogeveen, H., 10, 29, 129 Ireland, J. F., 12, 131 Johnson, S. L., 5, 237 Johnstone, R. A. W., 8, 151 Jonsall, G., 19, 223 Jose, S. M., 21, 197 Kemp, G., 20, 191 Kice, J. L., 17, 65 Kirby, A. J., 17, 183 Kohnstam, G., 5, 121 Kramer, G. M., 11, 177 Kreevoy, M. M., 6, 63; 16, 87 Kunitake, T., 17, 435 Ledwith, A,, 13, 155 Liler, M.. 11, 267 Long, F. A., 1, 1 Maccoll. A., 3, 91 Mandolini, L., 22, 1 McWeeny, R., 4, 73 Melander, L., 10, 1 Mile, B., 8, 1 Miller, S. I., 6, 185 340

Modena, G., 9, 185 More O’Ferrall, R. A., 5, 331 Morsi, S. E., 15, 63 Neta, P., 12, 223 Norman, R. 0. C., 5, 33 Nyberg, K., 12, 1 Olah, G . A., 4, 305 Page, M. I., 23, 165 Parker, A. J., 5, 173 Parker, V. D., 19, 131; 20, 55 Peel, T. E., 9, 1 Perkampus, H. H., 4, 195 Perkins, M. J., 17, 1 Pittman, C. U. Jr., 4, 305 Pletcher, D., 10, 155 Pross, A., 14, 69; 21, 99 Ramirez, F., 9, 25 Rappoport, Z., 7, 1 Reeves, L. W., 3, 187 Reinhoudt, D. N., 17,279 Ridd, J. H., 16, 1 Riveros. J. M., 21, 197 Roberston, J. M., 1, 203 Rosenthal, S. N., 13,279 Russell, G. A., 23,271 Samuel, D., 3, 123 Sanchez, M. de N. de M., 21, 37 Savelli, G., 22, 213 Schaleger, L. L., 1, 1 Scheraga, H. A., 6, 103 Schleyer, P. von R., 14, 1 Schmidt, S. P., 18,187 Schuster, G . B., 18, 187; 22, 311 Scorrano, G., 13, 83 Shatenshtein, A. I., 1, 156 Shine, H. J., 13, 155 Shinkai, S., 17, 435 Siehl, H.-U., 23, 63 Silver, B. L., 3, 123

CUMULATIVE LIST OF AUTHORS

Simonyi, M., 9, 127 Stock, L. M., 1, 35 Symons, M . C. R., 1, 284 Takashima, K., 21, 197 Tedder, J. M., 16, 51 Thomas, A., 8, I Thomas, J. M., 15, 63 Tonellato, U., 9, 185 Toullec, J., 18, 1 Tudos, F., 9, 127

Turner, D. W., 4, 31 Turro, N. J., 20, I Ugi, I., 9, 25 Walton, J. C., 16, 51 Ward, B., 8, I Westheimer, F. H., 21, 1 Whalley, E., 2, 93 Williams, D. L. H., 19, 38 1 Williams, J. M. Jr., 6, 63

341

Williams, J. O., 16, 159 Williamson, D. G., 1, 365 Wilson, H., 14, 133 Wolf, A. P., 2, 201 Wyatt, P. A. H., 12, 131 Zimmt, M . B., 20, 1 Zollinger, H., 2, 163 Zuman, P., 5, 1

Cumulative Index of Titles Abstraction, hydrogen atom, from 0-H bonds, 9, 127 Acid solutions, strong, spectroscopic observation of alkylcarbonium ions in, 4, 305 Acid-base properties of electronically excited states of organic molecules, 12, 131 Acids and bases, oxygen and nitrogen in aqueous solution, mechanisms of proton transfer between, 22, 113 Acids, reactions of aliphatic diazo compounds with, 5, 331 Acids, strong aqueous, protonation and solvation in, 13, 83 Activation, entropies of, and mechanisms of reactions in solution, 1, 1 Activation, heat capacities of, and their uses in mechanistic studies, 5, 121 Activation, volumes of, use for determining reaction mechanisms, 2, 93 Addition reactions, gas-phase radical, directive effects in, 16, 51 Aliphatic diazo compounds, reactions with acids, 5, 33 1 Alkylcarbonium ions, spectroscopic observation in strong acid solutions, 4, 305 Ambident conjugated systems, alternative protonation sites in, 11, 267 Ammonia, liquid, isotope exchange reactions of organic compounds in 1, 156 Antibiotics, p-lactam, the mechanisms of reactions of, 23, 165 Aqueous mixtures, kinetics of organic reactions in water and, 14, 203 Aromatic photosubstitution, nucleophilic, 11, 225 Aromatic substitution, a quantitative treatment of directive effects in, 1, 35 Aromatic substitution reactions, hydrogen isotope effects in, 2, 163 Aromatic systems, planar and non-planar, 1, 203 Aryl halides and related compounds, photochemistry of, 20, 191 Arynes, mechanisms of formation and reactions at high temperatures, 6, 1 A-S,2 reactions, developments in the study of, 6, 63 Base catalysis, general, of ester hydrolysis and related reactions, 5, 237 Basicity of unsaturated compounds, 4, 195 Bimolecular substitution reactions in protic and dipolar aprotic solvents, 5, 173 "C N.M.R. spectroscopy in macromolecular systems of biochemical interest, 13,279 Carbene chemistry, structure and mechanism in, 7, 163 Carbenes having aryl substituents, structure and reactivity of, 22, 31 1 Carbanion reactions, ion-pairing effects in 15, 153 Carbocation rearrangements, degenerate, 19, 223 Carbon atoms, energetic, reactions with organic compounds, 3, 201 Carbon monoxide, reactivity of carbonium ions towards, 10, 29 Carbonium ions (alkyl), spectroscopic observation in strong acid solutions, 4, 305 Carbonium ions, gaseous, from the decay of tritiated molecules, 8, 79 Carbonium ions, photochemistry of, 10, 129 Carbonium ions, reactivity towards carbon monoxide, 10, 29 Carbonyl compounds, reversible hydration of, 4, 1 Carbonyl compounds, simple, enolisation and related reactions of, 18, 1 342

CUMULATIVE INDEX OF TITLES

343

Carboxylic acids, tetrahedral intermediates derived from, spectroscopic detection and investigation of their properties, 21, 37 Catalysis by micelles, membranes and other aqueous aggregates as models of enzyme action, 17, 435 Catalysis, enzymatic, physical organic model systems and the problem of, 11, 1 Catalysis, general base and nucleophilic, of ester hydrolysis and related reactions, 5, 237 Catalysis, micellar, in organic reactions; kinetic and mechanistic implications, 8, 27 1 Catalysis, phase-transfer by quaternary ammonium salts, 15, 267 Cation radicals in solution, formation, properties and reactions of, 13, 155 Cation radicals, organic, in solution, kinetics and mechanisms of reaction of, 20, 55 Cations, vinyl, 9, 135 Chain molecules, intramolecular reactions of, 22, 1 Chain processes, free radical, in aliphatic systems involving an electron transfer reaction, 23, 271 Charge density-N.M.R. chemical shift correlations in organic ions, 11, 125 Chemically induced dynamic nuclear spin polarization and its applications, 10, 53 Chemiluminescence of organic compounds, 18, 187 CIDNP and its applications, 10, 53 Conduction, electrical, in organic solids, 16, 159 Configuration mixing model: a general approach to organic reactivity, 21, 99 Conformations of polypeptides, calculations of, 6, 103 Conjugated, molecules, reactivity indices, in, 4, 73 Crown-ether complexes, stability and reactivity of, 17, 279 D,O-H,O mixtures, protolytic processes in, 7, 259 Degenerate carbocation rearrangements, 19, 223 Diazo compounds, aliphatic, reactions with acids, 5, 331 Diffusion control and pre-association in nitrosation, nitration, and halogenation, 16, 1

Dimethyl sulphoxide, physical organic chemistry of reactions, in, 14, 133 Dipolar aprotic and protic solvents, rates of bimolecular substitution reactions in, 5, I73 Directive effects in aromatic substitution, a quantitative treatment of, 1, 35 Directive effects in gas-phase radical addition reactions, 16, 51 Discovery of the mechanisms of enzyme action, 1947-1963, 21, 1 Displacement reactions, gas-phase nucleophilic, 21, 197 Effective molarities of intramolecular reactions, 17, 183 Electrical conduction in organic solids, 16, 159 Electrochemical methods, study of reactive intermediates by, 19, I31 Electrochemistry, organic, structure and mechanism in, 12, 1 Electrode processes, physical parameters for the control of, 10, 155 Electron spin resonance, identification of organic free radicals by, 1, 284 Electron spin resonance studies of short-lived organic radicals, 5, 23 Electron-transfer reaction, free radical chain processes in aliphatic systems involving an, 23, 271 Electron-transfer reactions in organic chemistry, 18, 79 Electronically excited molecules, structure of, 1, 365 Electronically excited states of organic molecules, acid-base properties of, 12, 13 1

344


Energetic tritium and carbon atoms, reactions of, with organic compounds, 2, 201 Enolisation of simple carbonyl compounds and related reactions, 18, 1 Entropies of activation and mechanisms of reactions in solution, 1, 1 Enzymatic catalysis, physical organic model systems and the problem of, 11, 1 Enzyme action, catalysis by micelles, membranes and other aqueous aggregates as models of, 17, 435 Enzyme action, discovery of the mechanisms of, 1947-1963, 21, 1 Equilibrating systems, isotope effects on nmr spectra of, 23, 63 Equilibrium constants, N.M.R. measurements of, as a function of temperature, 3, 187 Ester hydrolysis, general base and nucleophilic catalysis, 5, 237 Exchange reactions, hydrogen isotope, of organic compounds in liquid ammonia, 1, 156 Exchange reactions, oxygen isotope, of organic compounds, 2, 123 Excited complexes, chemistry of, 19, 1 Excited molecules, structure of electronically, 1, 365 Force-field methods, calculation of molecular structure and energy by, 13, 1 Free radical chain processes in aliphatic systenls involving an electron-transfer reaction, 23, 271 Free radicals, identification by electron spin resonance, 1, 284 Free radicals and their reactions at low temperature using a rotating cryostat, study of8, 1 Gaseous carbonium ions from the decay of tritiated molecules, 8, 79 Gas-phase heterolysis, 3, 91 Gas-phase nucleophilic displacement reactions, 21, 197 Gas-phase pyrolysis of small-ring hydrocarbons, 4, 147 General base and nucleophilic catalysis of ester hydrolysis and related reactions, 5, 237 H,O-D,O mixtures, protolytic processes in, 7,259 Halogenation, nitrosation, and nitration, diffusion control and pre-association in, 16, I Halides, aryl, and related compounds, photochemistry of, 20, 191 Heat capacities of activation and their uses in mechanistic studies, 5, 121 Heterolysis, gas-phase, 3, 91 Hydrated electrons, reactions of, with organic compounds, 7, 115 Hydration, reversible, of carbonyl compounds, 4, I Hydrocarbons, small-ring, gas-phase pyrolysis of, 4, 147 Hydrogen atom abstraction from G H bonds, 9, 127 Hydrogen isotope effects in aromatic substitution reactions, 2, 163 Hydrogen isotope exchange reactions of organic compounds in liquid ammonia, 1, 156 Hydrolysis, ester, and related reactions, general base and nucleophilic catalysis of, 5, 237 Intermediates, reactive, study of, by electrochemical methods, 19, 131 Intermediates, tetrahedral, derived from carboxylic acids, spectroscopic detection and investigation of their properties, 21, 37 Intramolecular reactions, effective molarities for, 17,183


345

Intramolecular reactions of chain molecules, 22, I Ionization potentials, 4, 31 Ion-pairing effects in carbanion reactions, 15, 153 Ions, organic, charge density-N.M .R.chemical shift correlations, 11, 125 Isomerization, permutational, of pentavalent phosphorus compounds, 9, 25 Isotope effects, hydrogen, in aromatic substitution reactions, 2, 163 Isotope effects, magnetic, magnetic field effects and, on the products of organic reactions, 20, 1 Isotope effects on nmr spectra of equilibrating systems, 23, 63 Isotope effects, steric, experiments on the nature of, 10, 1 Isotope exchange reactions, hydrogen, of organic compounds in liquid ammonia, 1, 150 Isotope exchange reactions, oxygen, of organic compounds, 3, 123 Isotopes and organic reaction mechanisms, 2, 1 Kinetics and mechanisms of reactions of organic cation radicals in solution, 20, 55 Kinetics, reaction, polarography and, 5, 1 Kinetics of organic reactions in water and aqueous mixtures, 14, 203 p-Lactam antibiotics, the mechanisms of reactions of, 23, 165 Least nuclear motion, principle of, 15, 1 Macromolecular systems of biochemical interest, 13CN.M.R. spectroscopy in 13,279 Magnetic field and magnetic isotope effects on the products of organic reactions, 20, 1 Mass spectrometry, mechanisms and structure in: a comparison with other chemical processes, 8, 152 Mechanism and structure in carbene chemistry, 7, 153 Mechanism and structure in mass spectrometry: a comparison with other chemical processes, 8, 152 Mechanism and structure in organic electrochemistry, 12, 1 Mechanisms and reactivity in reactions of organic oxyacids of sulphur and their anhydrides, 17, 65 Mechanisms, nitrosation, 19, 381 Mechanisms of proton transfer between oxygen and nitrogen acids and bases in aqueous solution, 22, 113 Mechanisms, organic reaction, isotopes and, 2, 1 Mechanisms of reaction in solution, entropies of activation and, 1, 1 Mechanisms of reactions of p-lactam antibiotics. 23, 165 Mechanisms of solvolytic reactions, medium effects on the rates and, 14, 10 Mechanistic applications of the reactivity-selectivity principle, 14, 69 Mechanistic studies, heat capacities of activation and their use, 5, 121 Medium effects on the rates and mechanisms of solvolytic reactions, 14, 1 Meisenheimer complexes, 7, 21 1 Metal complexes, the nucleophilicity of towards organic molecules, 23, I Methyl transfer reactions, 16, 87 Micellar catalysis in organic reactions: kinetic and mechanistic implications, 8, 27 1 Micelles, aqueous, and similar assemblies, organic reactivity in, 22, 21 3 Micelles, membranes and other aqueous aggregates, catalysis by, as models of enzyme action, 17, 435 Molecular structure and energy, calculation of, by force-field methods, 13, 1

346


Nitration, nitrosation, and halogenation, diffusion control and pre-association in, 16, 1 Nitrosation mechanisms, 19, 381 Nitrosation, nitration, and halogenation, diffusion control and pre-association in, 16, 1

N.M.R. chemical shift+harge density correlations, 11, 125 N.M.R. measurements of reaction velocities and equilibrium constants as a function of temperature, 3, 187 N.M.R. spectra of equilibrating systems, isotope effects on, 23, 63 N.M.R. spectroscopy, 13C, in macromolecular systems of biochemical interest, 13, 279 Non-planar and planar aromatic systems, 1, 203 Norbornyl cation: reappraisal of structure, 11, 179 Nuclear magnetic relaxation, recent problems and progress, 16, 239 Nuclear magnetic resonance, see N.M.R. Nuclear motion, principle of least, 15, I Nucleophilic aromatic photosubstitution, 11, 225 Nucleophilic catalysis of ester hydrolysis and related reactions, 5, 237 Nucleophilic displacement reactions, gas-phase, 21, 197 Nucleophilicity of metal complexes towards organic molecules, 23, 1 Nucleophilic vinylic substitution, 7, 1 OH-bonds, hydrogen atom abstraction from, 9, 127 Oxyacids of sulphur and their anhydrides, mechanisms and reactivity in reactions of organic, 17, 65 Oxygen isotope exchange reactions of organic compounds, 3, 123 Permutational isomerization of pentavalent phosphorus compounds, 9, 25 Phase-transfer catalysis by quaternary ammonium salts, 15, 267 Phosphorus compounds, pentavalent, turnstile rearrangement and pseudorotation in permutational isomerization, 9, 25 Photochemistry of aryl halides and related compounds, 20, 191 Photochemistry of carbonium ions, 9, 129 Photosubstitution, nucleophilic aromatic, 11, 225 Planar and non-planar aromatic systems, 1, 203 Polarizability, molecular refractivity and, 3, 1 Polarography and reaction kinetics, 5, 1 Polypeptides, calculations of conformations of, 6, 103 Pre-association, diffusion control and, in nitrosation, nitration, and halogenation, 16. 1

Products of organic reactions, magnetic field and magnetic isotope effects on, 30, 1 Protic and dipolar aprotic solvents, rates of bimolecular substitution reactions in, 5, I73 Protolytic processes in H,O-D,O mixtures, 7,259 Protonation and solvation in strong aqueous acids, 13,83 Protonation sites in ambident conjugated systems, 11, 267 Proton transfer between oxygen and nitrogen acids and bases in aqueous solution, mechanisms of, 22, 113 Pseudorotation in isomerization of pentavalent phosphorus compounds, 9, 25 Pyrolysis, gas-phase, of small-ring hydrocarbons, 4, 147


347

Radiation techniques, application to the study of organic radicals, 12, 223 Radical addition reactions, gas-phase, directive effects in, 16, 5 1 Radicals, cation in solution, formation, properties and reactions of, 13, 155 Radicals, organic application of radiation techniques, 12, 223 Radicals, organic cation, in solution kinetics and mechanisms of reaction of, 20, 55 Radicals, organic free, identification by electron spin resonance, 1, 284 Radicals, short-lived organic, electron spin resonance studies of, 5, 53 Rates and mechanisms of solvolytic reactions, medium effects on, 14, 1 Reaction kinetics, polarography and, 5, 1 Reaction mechanisms, use of volumes of activation for determining, 2, 93 Reaction mechanisms in solution, entropies of activation and, 1, 1 Reaction velocities and equilibrium constants, N.M.R. measurements of, as a function of temperature, 3, 187 Reactions of hydrated electrons with organic compounds, 7, 115 Reactions in dimethyl sulphoxide, physical organic chemistry of, 14, 133 Reactive intermediates, study of, by electrochemical methods, 19, 13 1 Reactivity indices in conjugated molecules, 4, 73 Reactivity, organic, a general approach to: the configuration mixing model, 21, 99 Reactivity-selectivity principle and its mechanistic applications, 14, 69 Rearrangements, degenerate carbocation, 19, 223 Refractivity, molecular, and polarizability, 3, 1 Relaxation, nuclear magnetic, recent problems and progress, 16, 239 Short-lived organic radicals, electron spin resonance studies of, 5, 53 Small-ring hydrocarbons, gas-phase pyrolysis of, 4, 147 Solid-state chemistry, topochemical phenomena in, 15, 63 Solids, organic, electrical conduction in, 16, 159 Solutions, reactions in, entropies of activation and mechanisms, 1, 1 Solvation and protonation in strong aqueous acids, 13, 83 Solvents, protic and dipolar aprotic, rates of bimolecular substitution-reactions in, 5, 173 Solvolytic reactions, medium effects on the rates and mechanisms of, 14, 1 Spectroscopic detection of tetrahedral intermediates derived from carboxylic acids and the investigation of their properties, 21, 37 Spectroscopic observations of alkylcarbonium ions in strong acid solutions, 4, 305 Spectroscopy, "C N.M.R., in macromolecular systems of biochemical interest, 13, 279 Spin trapping, 17, 1 Stability and reactivity of crown-ether complexes, 17, 279 Stereoselection in elementary steps of organic reactions, 6, 185 Steric isotope effects, experiments on the nature of, 10, I Structure and mechanisms in carbene chemistry, 7, 153 Structure and mechanism in organic electrochemistry, 12, 1 Structure and reactivity of carbenes having aryl substituents, 22, 31 1 Structure of electronically excited molecules, 1, 365 Substitution, aromatic, a quantitative treatment of directive effects in, 1, 35 Substitution, nucleophilic vinylic, 7, 1 Substitution reactions, aromatic, hydrogen isotope effects in, 2, 163 Substitution reactions, bimolecular, in protic and dipolar aprotic solvents, 5, 173

348


Sulphur, organic oxyacids of, and their anhydrides, mechanisms and reactivity in reactions of, 17, 65 Superacid systems, 9, 1 Temperature, N.M.R.measurements of reaction velocities and equilibrium constants as a function of, 3, 187 Tetrahedral intermediates derived from carboxylic acids, spectrosopic detection and the investigation of their properties, 21, 37 Topochemical phenomena in solid-state chemistry, 15, 63 Tritiated molecules, gaseous carbonium ions from the decay of 8, 79 Tritium atoms, energetic, reactions with organic compounds, 2, 201 Turnstile rearrangements in isomerization of pentavalent phosphorus compounds, 9, 25 Unsaturated compounds, basicity of, 4, 195 Vinyl cations, 9, 185 Vinylic substitution, nucleophilic, 7, 1 Volumes of activation, use of, for determining reaction mechanisms, 2, 93 Water and aqueous mixtures, kinetics of organic reactions in, 14, 203

Advances in Physical Organic Chemistry

Advances in Physical Organic Chemistry

Advances in Physical Organic Chemistry. Volume 20